(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Annual report : National Institutes of Health. Division of Computer Research and Technology"



7/ 



! NATIONAL !NST!TUT£S OF HEALTH { 



NiH LiBRARY 



AUG 5 2GJG I 




BLDG 10, 10 CENTER DR. 
BETHESDA, 1^0 20892-1150 



PUBLIC HEALTH SERVICE - NATIONAL INSTITUTES OF HEALTH 
DIVISION OF COJMPUTER RESEARCH AND TECHNOLOGY 

' ANNUAL REPORTS 

Vol. II 

FY 1972-7i+ 



/_ . Z///''/^''-// -/. 



PUBLIC HEALTH SERVICE - NATIONAL INSTITUTES OF HEALTH 
DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 
Report of Program Activities 
July 1, 1971 through June 30, 1972. 



ANNUAL REPORT 
TABLE OF CONTENTS 

Page 

DIRECTOR' S SUMMARY 1 

Research Project Reports 

1.1 Visual and Biological Shape 7 

1.2 Clinical Center Admissions 9 

1.3 Computer Acquisition of Physician Orders 11 

1.4 Automated Processing of Medical Language 13 

1.5 Medical Journal Storage and Retrieval 15 

1.6 Information Processing System for Pathology Data 16 

1.7 "Computers in Clinical Medicine" Course 18 

COMPUTER CENTER BRANCH 

Summary 21 

Research Project Reports 

2.1 User Support and Communications 27 

2.2 Computer Center Users Guide 29 

2.3 HASP/Shared Spool 31 

2.4 360 Systems Development 33 

2.5 WYLBUR Text-editing System 35 

2.6 DATASTOR Program 38 

2.7 Lavender Box 40 

2.8 INTERFACE (technical communications) 43 

2.9 IBM 2550 Graphic Program Support 45 

2.10 CPS-Conversational Programming System 47 

2.11 PDP-10 System 50 



Page 

2.12 Computer Training Program 53 

2.13 On-line File Processing System 55 

2.14 Optical Document Processing System 57 

2.15 Chemical Structure Search and Display 59 

2.16 Macro Molecular Display and Manipulation 61 

2.17 Dial-Up Graphics 62 

2.18 Computer Output Microfilm 64 

2.19 Installation of Time Sharing Option (TSO) of 

OS/ 360 MVT 66 

LABORATORY OF APPLIED STUDIES 

Summary 69 

Research Project Reports 

3.1 Biomathematics and Statistics 73 

3.2 Evaluation of Computer-assisted Systems for Patient 

Care 75 

3.3 Computer Methods in Cardiologic Research 78 

3.4 Statistical Research in Clinical Pathology 82 

3.5 Applied Mathematics Unit 84 

3.6 Simulation and Modelling in Neurophysiological Research 86 
COMPUTER SYSTEMS LABORATORY 

Q1 

Summary ~ ~ ^^ 

Research Project Reports 

4.1 Computer Systems for NIDR 95 

4.2 Computer System for NIAMD 9 7 

4.3 Mass Spectrometer Related Work 99 



Page 
4. A Computer System for Gerontology Research Center 101 

4.5 Pulse Nuclear Magnetic Resonance (NMR) Spectroscopy 

for Carbon-13 and Proton Studies 102 

4.6 Computer System for NIMH 104 

4.7 Medical Telecommunications 105 

4.8 Cardiovascular Studies 107 

4.10 Computer Systems Laboratory Consultation 109 

4.11 Intensive Care Unit Computer System 110 

PHYSICAL SCIENCES LABORATORY 

Summary 111 

Research Project Reports 

5.1 Theory of Biochemical Separation Techniques 113 

5.2 Theory of the Helix-Coil Transformation of 
Polypeptides in Solution 115 

5.3 Molecular Mechanics 117 

5.4 Biophysical Analysis 119 

5.5 Carbon-13 Nuclear Magnetic Resonance Studies of 
Peptides and Proteins, including the use of 

Selective C^^ Enrichment 121 



5.6 Excitation and Transport Properties of Fluids; 

Laser Scattering 123 

5.7 Intermolecular Forces in Biological Structures 125 

5.8 Consulting Services 128 

5.9 Fundamental Studies 130 

5.10 Nuclear Magnetic Resonance (NMR) Studies of Proteins — 133 

5.11 High-Resolution Carbon-13 Fourier Transform NMR 
Spectroscopy at 55 MHz 137 



Pa£e 
HEURISTICS LABORATORY 

Summary 139 

Research Project Reports 

6.1 Automatic Question-Answering and Problem Solving 141 

6.2 Storage and Retrieval Studies — 144 

6.3 Automatic Pattern Recognition 146 

6.4 Chemical Notation and On-Line Information Systems 148 

6 . 5 Modelab 150 

6.6 Theoretical Applications to Programming 152 

6.7 Radiation Treatment Planning 154 

6.80 Information Storage and Retrieval with Index 156 

6.81 Microbiology Data Bank 158 

6.90 Nuclear Magnetic Resonance 160 

6.91 Mass Spectrometry Data Retrieval System 163 

6.92 Energy State Calculations 165 

6.93 Retrospective CBAC Literature Search 16 7 

DATA MANAGEMENT BRANCH 

Summary •'-6" 

Research Project Reports 

7.1 Cigarette Condensate Study (Phase 2) 171 

7.2 Carcinogenesis Bioassay Data System Implementation 173 

7.3 NIAID Research Contracts 176 

7.4 Collection of Bleeding Data Related to the Study of Kuru-178 

7.5 Consulting and Tutorial Services 180 



Page 

7.6 Bioassay Data Collection System 182 

7.7 Characterization of Normal and Abnormal Bladder 
Epithelium by Digital Computer 184 

7.8 Generated Update Facility 186 

7.9 Recursive Macro Actuated Generator 188 

7.10 Current Awareness Search of Chemical Literature 190 

7.11 Genetics of Lipoprotein Metabolism 192 

7.12 A Decision-theoretic Approach to Optimizing 
Radiotherapy 194 

7.13 A Clustering Algorithm for Biomedical Applications 197 

7.14 Cancer Survival System Processing of Dermatoglyphic 

Data 200 

7.15 NIH Space 202 

7.16 Generated File Reformatting Facility (REFORMATGEN) 204 

7.17 Framingham Heart Study, Exam 10 Inventory 206 

7.18 Leukocyte Infusion Update and Reporting System 208 

7.19 Generated File Reformatting Facility (TRANSACTGEN) 210 

7.20 Re-assigning Parking Permit Numbers when a State 

Issues New License Tags 212 

7.21 Type 11 Coronary Intervention Study 214 

7.22 Railroad Retirerr^ent 217 

7.23 URBS Systems Study 219 

7.24 Histocompatibility Computer Analysis 221 

7.25 Generated GPS Prompt Program 223 

7.26 Emergency Virus Isolation Facility Medical 

Monitoring Project 225 

7.27 Quick Report Generator 227 



Page 

7.28 Trends in Graduate Enrollment and Ph.D. Output in 
selected Science and Health Professional Fields 229 

7.29 Cancer Survival System 231 

7.30 NIH International Activitiies 233 

7.31 Medical Records: Diagnostic and Discharge 235 

7.32 Hepatitis Australian Antigen File Study 237 

7.33 Case Reports 239 

7.34 Nursing Education Study 241 

7.35 Data Retrieval from the Beta File 244 

7.36 Vitamin C Study 246 

7.37 Documentation Standards 248 

7.38 NICHD Patient Maintenance and Retrieval System 250 

7.39 C.C. Patient Register and Plate Making 252 

7.40 Biologies Controls Systems 254 

7.41 Ward Value Orientation Study 256 

7.42 Family Psychophysiology Project 258 

7.43 Animal Genetics 260 

7.44 Interview Scheduling System 262 

7.45 Math/Stat Library 265 

7.46 Nutrition Study of Laboratory Animals 268 

7.47 Chemistry Data Handling and Evaluation System 270 

7.48 Dyslipidemia Study 272 

7.49 M. Hyoiner Arthritis Study 274 

7.50 NIH Library Operations 276 

7.51 Collection and Analysis of Genetic Data 278 



Page 

7.52 Transposing WYLBUR Data into a Format Compatible 
with Publication Requirements of the Information 
(Including Heading, Footings, Page Numbers, etc.) 280 

7.53 Administrative Information Retrieval System 282 

7.54 NMR (Nuclear Magnetic Resonance) Literature Retrieval 
System 284 

7.55 NHLI Contract Grant System 286 

7.56 Tape Conversions and Plate Making 288 

7.57 Nursing Activities Study-13 East and Retrieval System 290 

7.58 Opportunity Skills System 291 

7.59 Printing Pressmen Study 293 

7.60 Clinical Laboratory Beta File Conversion 295 

7.61 Commercial Timesharing Users Survey 297 

7.62 Computer Analysis of Radioligand Assay and 
Radioimmunoassay data 299 

7.63 Cutting Oil Mist Study 301 

7.64 Surgery Branch Edit and Update System 303 

7.65 Aortic Valvular Disease Study 305 

7.66 Low Birth Weight for Age Study 307 



Serial No. 1.1 

1. Office of the Director 

2. 

3. Bethesda 

PHS - NTH 

Individual Project Report 

July 1, 19 71 through June 30, 1972 

Project Title: Visual and Biological Shape 

Previous Serial Number: 1.1 

Principal Investigator: Harry Blum 

Other Investigators: Virgil Carlson (MHLP) ; Francine Frome (NIMH and 
University of Maryland) ; Lewis Lipkin (NINDS) 

Cooperating Unit: NIMH, Lab. of Psychology, Section on Perception 

NINDS, Perinatal Research Branch, Section on Pathology 

Man Years 

Total: 0.9 

Professional: 0.9 
Other: 0.0 

Project Description: 

Objectives: 

The overall objective of this project is to develop a formal description for 
biological shapes and apply it to the development of a psychology of shape 
and an elaboration of shape processes in visual physiology. In addition, 
to provide a tool for the description of biological objects and their growth. 

Methods Employed: 

Develop and formalize a geometry of shape using growth processes. Explore 
implications of such a geometry to the visual process. Review relevant 
areas of experimental work in psychology and physiology. Design new 
experiments aimed at validating or refuting theoretical work. In addition, 
to develop hand and computer methods for extracting these new geometric 
descriptions . 

Major Findings: 

A pilot psychophysical shape experiment validated the theory for visual 
processes under the condition tested. An extension to a full scale 
experiment has been undertaken. Plans for a new visual experiment under 
different conditions have been started. In addition, a "do-it-yourself" 
hand method and a computer program for extracting these descriptions have 
been developed. 



Significance to Biomedical Research and the Program of DCRT 

The theory contributes a fundamental insight into several biological areas. 
First, it proposes a morphological base for description of cells, organs 
and organisms. This will be needed for taxonomic and diagnostic purposes, 
both human and automatic. Second, it opens up the area of shape perception 
which is needed for teaching and diagnostic purposes. Third, it proposes 
new principles of central nervous system organi/^ation which may be 
important to the neurophysiology of vision and to general understanding 
of the CNS. This extension encompasses motor, as well as sensory 
organization. Fourth, it has implications in embryonic, and later, 
growth and development since the theoretical principles apply to the 
efficient coding of biological shape. The visual and the growth coding 
are shown to be inversions equally well treated. Thus, the work, if correct, 
has potential for opening up new and important areas of biology which 
have been resistant to investigation before. 

Proposed Course: 

Theoretical work will continue. Experimental work on shape psychophysics 
will continue. It is hoped that experiments in visual physiology will be 
started. Collaborative work for the application of this description to 
cell nucleii, skull growth and heart shape during motion is being explored. 

Honors and Awards: None 

Publications: None 



Serial No. 1.2 

1. Office of the Director 

2. Medical Information Science 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 19 71 through June 30, 19 72 

Project Title: Clinical Center Admissions 

Previous Serial Number: None 

Principal Investigator: William C. White 

Other Investigators: Marvin Katz; Jon A. Halverson 

Cooperating Units: Office of Clinical and Management Systems, C.C. 

Man Years 



Total: 


1.2 


Professional: 


1.2 


Other: 


0.0 



'reject Description: 

Objectives 

To develop an on-line system for collecting and editing Clinical Center 
admissions data and establishing a file of biographic data on CC patients. 

Background: 

Computer Admission Systems have been designed for the Clinical Center in 
the past, however none of these systems was on-line and interactive to 
permit computer editing and validation of the data as entered. The 
availability of a commercial program for interactive terminal transactions 
on the Central Computer facility prompted this latest admissions system 
design. 

Methods Employed: 

Use of the "Customer Information Central System" (CICS) allows the use of 
the PL/1 programming language as well as Assembly language programs. The 
programs were tested and modified to fit the requirements and convenience 
of the potential users in the Clinical Center. A registry flip was 
established including all machine readable information on all past Clinical 
Center admissions. This registry file with its Soundex equivalent will 
provide the key to the retrieval of Biographic and Admission data and 
future patient medical record files. 



FY- 72 Activity 

The Admissions System was tested and has been installed with two terminals 
at the Admissions Office in the Clinical Center. Admissions personnel 
have been instructed and have been using the terminal for collecting and 
printing admissions data since the middle of March. 

Significance: / 

The Admissions System should provide the base for the development of a 
computer stored Patient Record for use in patient care and clinical research. 

Proposed Course: 

The Admissions System will be expanded to include information about 
patient transfers and discharges. Several other useful conputer output 
lists will then be available, such as the patient census and discharge 
summaries. The start of an on-line patient record system will be 
initiated using the doctor's orders and drug administration data for a 
test system. 



10 



Serial No. 1.3 

1. Office of the Director 

2. 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 



Project Title: Computer Acquisition of Physician Orders 

Previous Serial Number: Same 

Principal Investigator: Robert D. Gilbert, M.D. 

Other Investigators: None 



Cooperating Units; 



National Institute of Arthritis and 

Metabolic Disease, Arthritis and Rheumatism Branch 

Department of Nursing, Clinical Center 



Man Years: 




Total: 


2.3 


Professional: 


2.3 


Other: 


0.0 



Project Description: 

Objectives: 

The ultimate objective of this project is to acquire in machine processable 
■ form those physician orders which determine and initiate essentially all of 
the diagnostic and therapeutic activities of the hospitalized patient. In 
particular, those orders pertaining to drug therapy are of interest. Such 
information provides the basis for ultimate analysis of drug efficacy, 
toxicity, and interaction, as well as the eventual reduction of drug 
administration errors. 

Methods Employed: 

A DCRT supported public text editing system called WYLBUR has been the means 
used to acquire the basic data. Data is entered via computer typewriter 
terminal at the nursing unit, and is transmitted by telephone to the DCRT 
Central Computing Facility. 

Major Findings: 

The system has been in full operation on a single nursing unit for over 
twelve months. Nursing response to the system has been unanimously favorable. 
At least seven documents of clinical use to nurses are being generated daily 



11 



c 

from input data. Achieving compatability with nursing has been the 
initial objective of the system and this has been accomplished. 

Proposed Course: 

Additional programr.iing is underway to produce patient files of ordered 
drugs. From such files a system is being developed for the recording of ^1 
actual drugs administered to the patient. This information will then be > 
compared to clinical and laboratory data in the evaluation of drug ' 

toxicity and efficacy. 

Honors and Awards: None 

Publications: None _ . ' 



i| 



i 



12 



Serial No. 1.4 

1. Office of the Director 

2. Medical Information Science 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 19 72 



Project Title: 
Previous Serial Number: 
Principal Investigator: 
Other Investigators: 
Man Years : 



Automated Processing of Medical Language 

4.9 

Milos Pacak 

George Dunham; William C. White 



Total: 


2.5 


Professional: 


2.5 


Other: 


0.0 



Project Description: 

Objectives: 

The major objective of the project was design of a new and more sophisticated 
system for automated processing of medical language based on the experience 
with the present SNOP encoder, and medical data base. 

Background: 

For the past several years, an effort has been underway at DCRT to develop 
an automated information retrieval system to handle pathology data which 
were manually encoded and made available to the medical scientist for 
research purposes. We have developed a system for automated encoding 
(indexing) of pathology diagnoses which became operational this year. The 
system includes the acquisition of textual information, the interrogation of 
a dictionary (SNOP) , a set of morphosemantic rules and a set of syntactic- 
semantic rules to allow for the identification of the information content 
of the input messages. These components must interact to insure that the 
meaning of input messages is preserved and captured in a data structure 
suitable for information storage and retrieval. 

FY-72 Activities: 

As the encoder became operational refinements to the program and its 
special dictionaries were continued. Encoding of the Clinical Center 
surgical pathology diagnoses for 19 71 is complete and for 1970 largely 
complete. 



13 



In connection with further improvements of automated morphosemantic 
segmentation of compound medical terms, we started to work on the 
development of procedures for the construction of a medical lexicon to 
be used in computer processing of medical language. A man-machine 
interactive lexical refinement algorithm was developed in this direction. 

The design of a formal semantic model for the interpretation of medical 
records is being tested. 

Future Efforts: 

1) Encoding of pathology diagnoses for the years prior to 1970 

2) Improvement of morphosemantic segmentation and interpretation of 
medical compound words as a part of the study of the structure of 
medical lexicon and medical microglossaries . 

3) Formalization of procedures for the semantic interpretation of 
medical records. 

4) Construction of computer oriented medical dictionaries and micro 
glossaries. 

Honors and Awards: None 

Publications: 



14 



Serial No. 1.5 

1. Office of the Director 

2. Medical Information Science 

3. Bethesda 

PHS - NIH 
Individual Project Report 
July 1, 19 71 through June 30, 19 72 

Project Title: Medical Journal Storage and Retrieval 

Previous Serial Number: None 

Principal Investigator: William C. White 

Cooperating Units: National Institute of Child Health and Human 

Development, Intramural Research 

Man Years: 

Total: .8 

Professional: .1 
Other: .7 

Project Description: 

Objectives: 

The future objective of this project is to provide a computerized library 
reference service for NICHD scientists. The service should be an on-line 
system with interactive search query and browsing capability. 

Methods Employed: 

The medical research information published in the last four years of 
"Pediatric Research" has been acquired in machine readable form through the 
WYLBUR text editing system on the central computer facility. The Journal 
articles have been proofread, corrected and saved on tape back-up files. 

Major Findings: 

Effort to date has been mainly on getting the data base into the computer. 

Proposed Course: 

It is proposed to investigate some proprietary on-line retrieval systems for 
possible use by NICHD through the central computer. One such system is an 
IBM program product that operates under the Customer Information Control 
System (CICS) . 



15 



Serial No. 1.6 

1. Office of the Director 

2. Medical Information Systems 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 19 72 

Project Title: An Information Processing System for Pathology Data 

Previous Serial Number: None 

Principal Investigator: Martin Epstein 

Other Investigators: Samuel Harper 

Karen Lantner 
Paul Kall:ov7ski 

NCI, Laboratory of Pathology 



Cooperating Unit: 


1 


Man Years: 




Total: 


2.8 


Professional: 


2.4 


Other: 


.4 



Project Description: 

Objectives : 

The objective of this project is to develop and maintain an infoirmation 
processing system for pathology data. This project includes the creation 
and updating of a date base for Autopsy, Surgical Pathology, and 
Cytopathology data, and retrieval and reporting from this data base 
through queries submitted by medical investigators. The Systematized 
Nomenclature of Pathology (SNOP) is maintained on disk for use in 
automated encoding and in the development of suitable encoding 
dictionaries for computer processing. Techniques are being developed to 
classify automatically medical diagnoses selected on retrieval from data 
base. 

Progress : 

During the past year a data base for surgical pathology data has been 
created which accepts as input data which have been automatically encoded. 
This permits data to be input at a terminal, encoded, a data file created, 
and retrieval to be performed on this file. This system allows the medical 
researcher to interrogate the data base and to display a subset of the 
file for further study. An initial program has been written to analyze 
the output from a retrieval run and to group the results using clustering 
techniques , 



16 



Proposed Course: 

Extension of the current research to include the processing of other 
medical data. Each of the data bases is to be created in a modular 
fashion to permit information to be merged across the data files as they 
become available. The study of algorithms useful in clustering data is 
continuing. An effort is underway to develop procedures to group 
related diagnoses and thus provide a feedback mechanism for creating 
encoding dictionaries, and to gain insight into efficient data structures 
needed to represent medical records. 



17 



Serial No. 1.7 

1. Office of the Director 

2. 

3. Bethesda 



PHS-NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: "Computers in Clinical Medicine" Course 

Previous Serial Number: None 

Project Leader: W.C. Mohler 

Other Investigators: G. Knott, R. Sproull, J. Prewitt, M. Epstein, S. Allen, 
K. Kempner 

Cooperating Units: (See Methods employed) 

Man Years: 

Total: 1.2 
Professional: 1.0 
Other: 0.2 

Project Description: 

Objectives: Develop and give an elective course about clinical 

application of computers for medical students as part of NIH 
Clinical Elective Program. 

Methods employed: The curriculum was developed in-house, largely from 
first principles, by consultation among those who were interested 
in the course. The resulting nine-week course contained about 120 
sessions given by a faculty of about 60 volunteers. There were 
site visits to applications in Washington and Baltimore, and guest 
lectures and panelists mostly from the same area. Computing was 
taught largely on PDP-10, with IBM 360/370 for some applications. 

Major Findings: The course as designed was acceptable to the students 
and provided a good learning experience for students and faculty. 
It did not, however, serve as a significant focus of interest or 
learning for NIH staff beyond those who participated as faculty. 

Significance to Biomedicine and NIH mission: The role of computer 

technology and information science in medicine is a much discussed 
but often confused subject. This course served as an experimental 
prototype to design and test various methods and materials which 
will be of value to medical educators elsewhere. 



18 



Proposed Course: The curriculum this year was based largely on 

objectives chosen by subject matter specialists in the faculty. 
Based on the experience from its first running, we will revise 
the course for the coming year to incorporate some of the objec- 
tives added by the students during their evaluation of the course 
and its parts. 



Honors and Awards : None 
Publications: None 



19 



July 1, 1971 thru June 50, 1972 

PUBLIC HEALTH SERVICE-NATIONAL INSTITUTES OF HEALTH 
DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 

1. DCRT-2 



Serial Number 
2. COMPUTER CENTER BRANCH 3. J. D. Maughton 



Branch Chief 



SUMMARY 



The Computer Center Branch designs, plans, implements and 
maintains a large general -purpose computer utility to meet most 
effectively the dynamic and diverse requirements of both ^'.I.H. 
research investigators and managers in the support of modern 
medicine. This charge Includes the development of unique system 
facilities that will bring the computer to bear on problems at 
every level of research and In many locations. The core of this 
computer facility Is a network of computers and remotely located 
terminals, which, by means of modern communications techniques, 
makes the power of the computer immediately available to all 
laboratories and offices throughout N.I.H. An inherent 
responsibility of the utility Is the continued research into 
extending the network Into the research environment and adapting 
to the constant impact of new knowledge and program direction. 

A full spectrum of computing power is provided to research 
personnel. These facilities Include conversational programming, 
graphics, microfilm output, text editing, remote job entry, time 
sharing and batch processing. Large systems as well as 
mi nl -computers and terminals are lashed together providing 
capability at many levels. Research into the computer and 
information sciences coupled with the N.I.H. medical 
investigators' needs and cooperation will bring computers closer 
to the research environment where they can perform most 
effectively in attacking the complex problems of modern medicine. 

The medical research programs of N.I.H. require the most powerful 
and flexible of computer services and tools available today. The 
computer network provided must have a distributive power that is 
easily accessible on demand to scientists In the laboratory 
itself. The goal is to mold, polish and, in general, enhance the 
computer into a complete tool for medical research and its 
administrative support. New areas of computer applications will 
be sought out continuously. In conjunction with a comprehensive 
program of educating new and old computer users In how to use 
computers most effectively in their domains. 



21 



1972 Activities 

The NIH computer utility continued its remarkable growth rate of 
25!!; while satisfying the demands for computational support by the 
MIH scientists/ researchers and their administrative support. 
The magnitude of the dramatic increase in workload is reflected 
by how much the computer network has grown. The number of remote 
batch terminals increased from 18 to 11, interactive typewriter 
terminals from 250 to over UOO, the workload in general grew 25% 
from 80,000 to over 105,000 jobs a month, while terminal sessions 
went from 1,000 to over 1,500 a day. History was made by the 
computer utility as the first IBM 370/165 was installed to meet 
the increasing demand for on-line daytime processing. This new 
system replaced both a 360/65 and a 360/50. Late In the year a 
second 360/65 was replaced by another 370/165. These major 
changes in the utility were necessary to Insure that response 
time remains adequate to meet the needs of the KM H research 
mission. The users of the utility benefited in two ways. The 
favorable response time helped their job turnaround, but more 
importantly improved efficiency allowed a 10% rate reduction on 
all computer services offered by the Center and an additional 20% 
reduction for certain types of overnight work. Thus the user 
continues to get more for his computer dollar from the utility. 
The use of and dependence upon on-line data bases mirrored the 
general growth pattern. Shared disk space expanded to 
approximately three billion bytes of storage to satisfy this 
need. 

The Installation of new advanced hardware was not the only v/ay 
the Computer Center net the challenge of a burgeoning workload. 
The computer operating systems are continuously tuned to remain 
sensitive and responsive to the research communities needs. The 
current system Is continuously studied; as weak points are 
Identified, ways to overcome the weaknesses are proposed, tried 
and installed. Similarly the latest In new computing techniques 
are studied. Those that are applicable to our environment are 
studied further and installed if there are significant benefits 
to be gained by their use. The common (shared) work queue Is an 
example of this process. Machine room organization is also 
studied. Our complete machine room was reorganized recently. 
This reorganization allows for a more effective and flexible 
operation. To compliment the computer room reorganization, a 
common or shared console addition to HASP has been designed and 
Is being programmed. When completed operations will have 
complete control of Its environment, from a single console. 

The growth of the workload can be attributed in part to the 
continuing effort of the Computer Center to educate the 
scientists and administrators in the use of Information science 
techniques. This educational process makes the computer user 
more aware of the possibilities of the utility and more 
productive In his use of the system. The number of students 

22 



attending the 80 courses offered in the computer training program 
passed 2000. Stilly many more applied for training, but could 
not be handled due to the lack of sufficient staff. 

The latest In computing and Computer Center activities Is 
broaricast to the entire computing community thru INTERFACF, a 
series of technical notes, and the Computer Center's Users Gui He . 
INTERFACE and the Users Gui de have a wide circulation/ and have 
been used as models for similar publications by many data 
processing installations. The quality of the INTERFACE sections 
is attested to by at least one other installation that asl<ed for 
and was given permission to publish, in their newsletter, 
complete sections of INTERFACE as written. 

The computer user, be he scientist or administrator, is an 
Important Individual. The staff of the User Services and 
Assistance Section gives Individual high quality help to all 
users, no matter what their problem. They insure the users of 
consistent bugfree systems software and keep the user up to date 
by the distribution and updating of all pertinent Computer Center 
material, technical and otherwise. Another important asset to 
the user community is the service provided by the Program Support 
Section. Thru this section a program execution and programming 
maintenance service is available to all t'lH organizations having 
recurring application systems. This group has worked with many 
users, large and small, relieving them of the burden of recurring 
systems by meeting established schedules, coordinating data flow, 
and updating the systems in time of hardware and software 
changes . 

August saw the graduation of the first class of students 
completing the Computer Center's Computer Operator Training 
Program. This program v;as a part of the educational, career and 
advancement opportunities for NIH employees under the DHEl/'s New 
Career Program. All students of the course have obtained 
full-time employment as computer operators in various 
Installations thruout the federal government. Speaking on 
effective mobility. Dr. Marston, Director of NIH, suggested that 
the computer operator training program might be a good one for 
others to examine since It embodied six factors essential to 
successful change In scientific programs and in programs to help 
employees achieve full potential in their jobs. 

To help those users producing large quantities of output a 
microfilm output device was acquired and put into general 
service. With microfilm as an output medium the user needs less 
space to store his output and his costs decrease. Further, he is 
given a choice of microfilm strip or microfiche cards. Thus his 
output can be stored, manipulate^ and processed faster, at a 
considerable savings of cost, and when linked to appropriate 
viewers, manpower. Data searches and retrieval become possible 



23 



with properly designed output. Microfilm becomes another general 
purpose service provided the scientific and administrative 
communi ty . 

The combination of computer techniques, biological applications, 
and collaboration with NIH scientists paved the way for vast 
Improvement in the techniques for the storage, search and display 
of chemical structures. A user oriented chemical Information 
system was the product of this work. An Instantaneous, 
retrospective search of chemical literature for occurrence of 
certain terms Is now possible. New structure files of interest 
to wider groups of scientists are being established. 
Particularly, preparations are under way for establishing the 
complete chemical abstracts (CAS) registry file of over 2 million 
compounds. The existence of this complete registry file will 
guarantee a scientist doing a chemical structure search that the 
search results are meaningful and complete. 

Substantial progress has been made in the area of processing 
on-line file systems. The resident monitor that provides the 
data base/data communications interface has been Implemented. 
The activities of the NIH Clinical Center Admissions Department 
was the pilot project used to test the system. The admissions 
system v/as successfully implemented and is now operational. 
Additional areas within the Clinical Center have been selected 
and systems for them are being designed and implemented to 
utilize the on-line data base system. 

The popularity of the interactive terminal systems CPS and WYLBUR 
continues to increase. The grovvth in the number of terminals 
(over UOO) attests to their popularity and utility. The number 
of telephone lines servicing these systems has grown to 
approximately 200 and there are plans for another hO in the 
upcoming months. The general acceptance of Interactive, as well 
as remote job entry, terminals is demonstrated by the fact that 
over 90?^ of our current wori<load Is now generated at terminals. 
With the growth of the number of terminal users it is difficult 
to pick out outstanding uses of the system. Many scientists and 
administrators are making effective use of their tools in many 
Ingenious ways to solve their own problems. The users Interest 
in and use of the terminal systems has prompted them to make many 
useful suggestions about improvements in the terminal systems. 
These suggestions along with our own observations led to a 
planned major revision and expansion of the WYLBUR system. The 
end result v/1 1 1 be a more efficient and more powerful tool than 
we currently have. CPS, meanv/hile, continues Its growth at a 
somewhat slower rate. Improvements to the system are a 
continuous process. Response time was decreased significantly 
when the system was installed on a faster processor, while 
overall performance and reliability has helped multiple system 
changes . 



24 



Looking to the future of interactive processing considerable work 
has already been done to provide a general time sharing system to 
the computing community. With a general purpose time sharing 
system the user v;I 1 1 be able to design and implement his own 
on-line interactive system. In any language of his choosing that 
he is familiar with. Time sharing offers some users decreased 
debugging time, thus systems become productive faster. Rut the 
ability to implement their own interactive system, communicating 
with the terminal In the laboratory promises to be a boon to many 
NIH scientists and administrators. When development of the time 
sharing system is complete and relatively free of errors the 
system will be offered to all as a service of the Center. 

The Investigation of the use of small (mini) computers as 
terminals or as part of a total system continues In many 
directions. The development of a dial-up computer terminal with 
graphics capability av/aits state-of-the-art advances in storage 
tubes to become really effective. A lash up between two computer 
systems, the AGT-30 and PDP-10, has led to the development of a 
macro molecular display system. Scientists can view and 
manipulate macro molecular models. Computer drawn diagrams of 
molecule sections for publication can also be generated. V/ork 
with time sharing monitors and hardware paging systems is In 
progress eventually offering users interactive systems on small 
computers. The use of mini computers to control and organize the 
growing communications needs of the Computer Center advanced 
significantly. A protocol for Intercomputer communications was 
established as v;el 1 as a message switching mechanism. This will 
enable any mini computer with teleprocessing capabilities to 
communicate with the central utility and eventually any other 
mini computer. Thus the provision of large computer facilities 
to laboratories with mini computers brings the complete force of 
large computers to bear on many laboratory projects. 



25 



Serial Mo. 2^ 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: User Support and Communications 

Previous Serial flumber: Same 

Principal Investigator: Frances E. Halverson 

Other Investigators: Staff of PAL Unit 

Cooperating Units: None 

Man Years: 

Total : 8 
Profess I onal : 8 
Others: 

Project Description: 

Obj ect I ves : 

To provide the users of the Computer Center with the 
personal assistance necessary if they are to make 
effective use of the Center's facilities. To provide 
users assistance in resolving problems encountered while 
using the systems (hardware and software) maintained by 
the Center. 

Methods Employed: 

The PAL (Programmer Assistance and Liaison) Unit was 
established to perform this function. Thru the Bugs, 
Diagnostics and Hints Section of the Computer Center's 
technical report, INTERFACE, and the Users Guide , the PAL 
Unit keeps the user community abreast of the latest word 
in problems, system changes and programming hints. They 
also assist the Systems Team In the design, selection 
installation, testing and modification of all system 
software provided by the Computer Center. Another method 
of vyritten communication Is the Programmer Trouble Report 
(PTR). The Unit processed 1700 PTR ' s in 1971. This form 



27 



allows the user to make complaints, suggestions or 
comments/ requests or to describe a problem he needs 
help with. The information from this form helps PAL and 
CCB to formulate future policies, plan systems changes 
and inform IBM or the Systems Team of software/hardware 
problems. Five of the eight staff members teach in 
the CCB educational program on a regular twice/year basis. 

Much of the time that is not spent working directly on 
users' programs is devoted to maintenance of the current 
system and on-line disk packs and to applying and testing 
vendor-supplied fixes to the system. They are also 
responsible for testing and documenting new releases of 
the operating system. This involves developing and 
running the NIH test job stream against each nev^ release 
to test its integrity. The test job stream, which is 
made up of jobs submitted by users, is intended to be 
representative of the work done at the MIH Computer 
Center. By evaluating the results of these runs, the PAL 
Unit is able to prepare users in advance for any 
necessary changes and to accomplish the transition from 
release to release with a minimum of inconvenience to the 
users. They evaluate new terminal systems and software 
packages being considered for implementation. 

Significance to Program of the Division: 

The computer user at NIH has at his disposal a group of 
competent professional programmers that are able to 
assist him with his problems In running programs or using 
the computer facility. The PAL Unit notes all trouble 
areas and, through INTERFACE and the PTR form, 
communicates common problem areas to all users. In all 
facets of computer use the user has someone that can give 
him the answer or guarantee to find the answer for him. 

Honors and Awards: None 

Publ i cat i ons : None 



28 



Serial No. 2^ 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - MIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 



Project Title: Computer Center Users Guide 

Previous Serial Mumber: Same 

Principal I nvest i ;;ator : Frances Halverson 

Other Investigators: Leslie BardeP/ PAL Unit and Systems Team 

Cooperating Units: None 

Man Years: 

Total: 1 
Prof essi onal : 1 
Others: 

Project Description: 

Obj ecti ves : 

To provide the users of the central computer facility with 
a guide to the services, standards and use of the Computer 
Center. 

Methods Employed: 

A 200-page Users Gui de was published and distributed to 
all users of the Computer Center, to Institute and Divi- 
sion Administrative or Executive Officers, to contract 
companies required to use the Computer Center and to 
other organizations and Individuals having a logical 
need for it. Updates to the Users Gul de are published 
and distributed as necessary to keep it current. 



29 



Significance to Program of the Division: 

For the first time all information pertaining to computing 
was brought together in a single reference document. The 
Users Gui de contains a description of all facilities and 
services and how to use them. Programming standards, 
languages supported, JCL summary and other facets of com- 
puting are all given in detail. It is a complete guide 
to computing for the computer user. 

Honors and Awards: None 

Publ i cat ions : None 



30 



Serial No. 2_.J 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1^ 1971 through June 50, 1072 

Project Title: HASP/Shared Spool 

Previous Serial Number: Same 

Principal Investigator: James Oberthaler 

Other Investigators: William Jones 

Cooperating Units: None 

Man Years 

Total: 1.6 
Prof essl onal : 1.6 
Others: 

Project Description: 

Background : 

The HASP system Is an extension to the principal control 
program employed on each of the Computer Center's IBM 
System/360 computers. HASP is responsible for reading 
all jobs into the multi CPU system, scheduling their 
execution on a priority basis, and handling all printed 
and punched output from the jobs. HASP provides the 
Interface whereby conversational terminal systems enter 
programs into the batch stream and control their 
execution. 

Obj ect 1 ves : 

1. To tailor the standard HASP distributed by IBM to the 
specific needs of the NIH computing community. 

2. To integrate the ^! I H modifications to HASP into the 
nev/ versions of HASP which are released periodically 
by IBM. 



31 



Methods Employed: 

HASP is a computer program written in IBM System/360 
Assembler Language. The standard version distributed by 
IBM consists of approximately 10,000 instructions to which 
NIH has now added on the order of an additional 10,000 
i nstructions. 

Major Findings: 

A complete rewrite of HASP by the IBM implementation team 
necessitated an overhaul of the entire package of 
modifications which NIH has made to HASP. This new 
version of HASP was installed at NIH during May 1972. 

Significance to Program of the Division: 

The HASP system is of prime importance to the entire 
batch processing system at CCB, because HASP is signi- 
ficantly easier to modify than the Operating System, and 
is not sensitive to changes made in the Operating System 
it contains many provisions that are unique to NIH. 

The multiple computer environment at the Computer Center 
allows significantly better overall service to the MIH 
computing community by providing both increased capa- 
bility and superior reliability to a single machine 
operation. V/ithin this framework, however, the task of 
managing three computers rather than one presents 
formidable scheduling and logistical problems. Through 
the use of a shared input/output queue, many of the 
decisions currently being made manually can be automated, 
thus providing more immediate and accurate response to 
the variations in work load and resource availability. 

Proposed Course: 

To develop an adjunct to the HASP Shared Spool which will 
enable operators on a given system to issue commands to 
be executed on another system. This will support a 
considerable physical reorganization of the machine room. 

Honors and Av/ards: None 

Publications: None 



32 



Serial No. 2jJi 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - MIH 

Individual Project Report 

July I, 1971 through June 30, 1972 

Project Title: 360 Systems Development 

Previous Serial Mumber: Same 

Principal Investigator: Robert H. Brunelle 

Other Investigators: Staff of Systems Team 

Cooperating Units: PAL Unit 

Man Years: 

Total: 6 
Professional : 6 
Others: 

Project Description: 

Obj ecti ves : 

To maximize the thruput and minimize turnaround time to 
all users of the central facility's 360 computers. To 
put computing power into every researcher's lab or office 
thru remote terminals and software systems. Develop the 
concept of the computer utility to its fullest. In 
general/ provide the MIM user with the best Computer 
Center and service he can get thru software development 
and hardware expansion. 

Methods Employed: 

Judicious selection and tailoring of software systems to 
the MIH environment. Acquisition and development of 
software and hardware to provide the researcher with the 
tools he needs. Constant attention to overall system 
software and hardware needs so that new systems and hard- 
ware are available when needed. Increasing the capacity 
of the system to keep ahead of the needs of the NIfl users, 



33 



Significance to Program of the Division: 

Constant attention to all facets of computing provides 
the MIH with an up-to-date computer facility catering to 
the needs of all MIH's researchers. Computer pov/er in 
the lab brings a powerful research tool closer to the 
project, thus making it easier to use, and more lii<ely to 
become an integral part of, the research program. Close 
attention to daily running permits us to tune the total 
system, maximizing throughput while minimizing turnaround 
time, thus remaining responsive at all times to the 
constantly growing workload. All of these efforts 
combine to form a forward looking Computer Center 
sensitive to the needs of the NIH research environment. 

Proposed Course: 

To monitor the NIH environment, continuing to mold the 
computer facility to best meet the needs of NIH. 

Honors and Awards: None 

Pub 1 i cat ions : None 



34 



Serial Mo. 2_iJ 

1. Computer Center Branch 

2. 

3. BethesHa 



PHS - MIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: WYLBUR Text-editing Systen 

Previous Serial Number: Same 

Principal Investigator: James Oberthaler 

Other Investigators: Leslie Barden 

Cooperating Units: None 

Man Years: 

Total : 0.6 
Profess ional : 0.6 
Others: 

Project Description: 

Obj ect i ves : 

1. To provide Computer Center users with a convenient 
and comprehensive tool to aid in the creation and 
development of computer programs and data files. 

2. To make it significantly easier to create and edit 
text materials such as letters, reports, etc. 

3. To reduce the necessity for users to leave their 
laboratories or offices in order to use the computer, 

k. To provide a convenient method for ascertaining the 
status of the computing system as a whole and locat- 
ing particular jobs as they are being processed. 

Methods Employed: 

17YLBUR is an on-line text-editing and remote job entry 
system oriented to low-speed character node terminals 
such as the IBM 27i4l, 1050, and 2260 and Models 33, 35 
and 37 teletype machines. It was originally designed 
and implemented at the Stanford University Computation 

35 



Center. WYLBUR provides the user, in his lab or office, 
the facility to create and edit source programs in real 
time, submit them for compilation and execution by the 
standard job stream processor, and retrieve the results 
of execution at his terminal. In addition, the user may 
create, store and modify text material, such as letters, 
reports, books, etc. WYLBUR is constantly being modified 
to Improve performance and make new facilities available 
to the users. 

WYLBUR consists of two computer programs written in IBM 
System/360 Assembler Language which reside permanently in 
two separate regions under OS/360. MILTEN, the tele- 
communications monitor, controls all input and output for 
the lov;-speed terminals and handles communication with 
the system operator. MILTEN can communicate with one or 
more subsystems operating in other regions under OS/360. 
All subsystems use the same pool of low-speed lines, thus 
reducing the cost and providing more flexibility for the 
user. WYLBUR, the text editor, is a subsystem of MILTEN. 
It performs all of the actual manipulation of text and 
the remote job entry to HASP. WYLBUR calls on MILTEN In 
a device Independent manner for all terminal I/O requests. 
The permanent text files are stored on 231tt disks which 
are shared among all CPU's in the system. This permits 
batch jobs to access WYLBUR data sets. The text is stored 
in a compressed format in order to save space. The work- 
ing files arp. stored on 2301 drum (with disk as backup) 
and are moved in and out of core under a demand paging 
al gor 1 thm. 

Major Findings: 

l/YLBUR has been extensively modified in the last year in 
order to increase its usefulness to users and to Increase 
1 ts capacl ty. 

1. The number of telephone lines into WYLBUR has been 
Increased from 92 to 111 in order to keep pace with the 
constantly Increasing demand. The maximum number of 
users simultaneously logged on (111) has been reached 

f requent 1 y. 

2. The number of online file disks has been Increased 
from 22 to 36 in order to satisfy the increasing demand 
for storage space. 

3. WYLBUR disk and batch stream communication modules 
have been completely rewritten to permit multi-path 
access to these facilities. This change was necessary 
to reduce a severe bottleneck which was encountered 
when 70 or more users were logged on simultaneously. 



36 



k. New options were added to the SHOVJ DSNAMES command 
to enable the user to monitor his permanent disk data 
sets more effectively. 

5. The CHANGE command was modified to enable the user 
to translate to upper case data which was inadvertently 
entered in lower case mode. 

6. Other commands and options have been modified in 
order to Increase their usefulness. 

7. Multiple paging files permit WYLBUR to make use of 
the additional temporary storage space needed to support 
many more users. 

In addition to the effort described above, there Is an 
extensive program to educate the users about WYLBUR. 
Many sessions of the WYLBUR for Programmers and the WYLBUR 
for Secretaries and Administrative Personnel courses 
have been taught/ and a WYLBUR Reference Card was 
pub! i shed. 

Significance to Program of the Division: 

VJYLBUR represents a significant move towards lowering 
the amount of non-programming overhead Involved In the 
development of programs. With the job output available 
at the user's terminal, effective turnaround time is 
substantially lowered. In addition, the added ease of 
using a typewriter-like terminal Instead of a keypunch 
contributes to more rapid development and debugging of 
programs. WYLBUR has also gained wide acceptance as a 
tool to aid In the preparation of memos, reports, and 
other documents. In addition, V/YLBUR has been used for 
data collection and Inquiry applications. 

Proposed Course: 

A total rewrite of WYLBUR Is being considered. It v;as 
designed origntally for 50 to 50 users. Since we have 
typically In excess of 100 simultaneous users, and when 
telephone limits are solved the number will quickly grov/ 
to 200, reoptlmlzatlon of WYLBUR Is a definite require- 
ment. 

Honors and Awards: None 

Pub 1 I cat I ons : None 



37 



Serial No. 2_»i 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 



Project Title: DATASTOR Program 

Previous Serial Number: Same 

Principal Investigator: James Oberthaler 

Other Investigators: William Jones 

Cooperating Units: None 

Man Months : 



Total : 


1/2 


Profess! onal : 


1/2 


Others: 






Project Description: 
Objecti ves : 

1. To provide specifications for a general purpose 
processor-to-processor communications facility. 

2. To implement, using these specifications, a set of 
computer programs which will permit computers remote from 
the main CCB facility to create, manipulate, and retrieve 
data files on the Computer Center's disk storage devices. 

Methods Employed: 

1. Communications links between the central and remote 
sites may utilize switched network or leased line 
facilities; both synchronous and asynchronous trans- 
mission techniques are supported. 

2. The programs which compose the DATASTOR system are 
coded in System/360 Assembler Language. 

3. The programs operate in the batch environment at 
present, as opposed to being a permanently on-line 
system, such as WYLBUR. 

38 



Major Findings: 

Tests have been performed to check out communications 
hardware from two UNI VAC DDP516 computers/ one at the 
Arthritis Institute (Building 2), the other at the Dental 
Insti tute (Bui Idi ng 30) . 

It Is anticipated that these machines will shortly become 
users of the DATASTOR system. 

Significance to Proeram of the Division: 

This Is an Important step In the Computer Center's plan 
to create an environment In which the many small 
specialized laboratory computers can easily converse with 
the powerful central facility. 

Proposed Course: 

1. To complete the Implementation of the system according 
to the specification developed. 

2. To develop a facility within the framework of the 
system for remote computers to submit batch jobs and 
receive their output. 

3. To institute permanently on-line service by the 
system, as soon as Its utilization Is sufficient to 
justify doing so. 

Honors and Awards: None 

Publ icat ions : None 



39 



Serial No. 2jJ 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 



Project Title: Lavender Box 

Previous Serial Number: None ^. ^ 

Principal Investigator: R. Sproull 

Other Investigators: R.H. Brunelle, W. Speary 

Cooperating Units: None 

Man Years: 

Total: 1/2 
Profess ional : 1/2 
Others : 



Project Description 
Obj ect i ves : 



The Lavender Box project Is intended to organize the 
growing communications needs of the Computer Center. The 
360/370 facility and the PDP-10 system offer many 
terminal services to users. The terminals are attached 
to the computer via telephone lines. As the use of 
terminals has increased, communications problems have 
become acute. The Lavender Box project is an attempt to 
solve some of these: 

1. Telephone issues. Organizing the sheer clutter of 
more than 200 telephone lines. Providing informative 
recorded messages when systems are unavailable or down. 




40 



3. Providing computer facilities to laboratory 
computers. Laboratory computers should be able to use 
the full facilities of the Computer Center via 
communications links. 

k. Load balancing - As the terminal usage grov/s, no one 
computer can be expected to service all users. Instead, 
Incoming requests for service should be able to be 
switched to available computers. 

Methods Employed: 

A specific design for the "Lavender Box" has been 
formulated (internal document: Initial Lavender Box 
Design, W. Speary & R. Sproull). The scheme involves two 
distinct concepts: (1) establishing a message-swi tch i n?; 
network which will allow any computer process to 
communicate easily with any other computer process, even 
though they may be miles apart, and (2) providing two or 
three sites on this network which will accept incoming 
calls from users. These sites will have special computer 
programs which satisfy user's computation requests by 
communicating with processes in the network which can 
perform the desired computation. 

The design is not complete, but suggests a philosophy 
of construction of the system. An RFP is being prepared 
to process a detailed design, together with estimates of 
hardware requirements. 

Major Findings: 

The small, limited objective network we propose Is 
the result of a compelling need to distribute computation 
among a variety of machines; each satisfying requests 
according to its load and capabilities. This principle 
has already been applied to batch processing at the NIH 
Computer Center. The Lavender Box represents an 
extension of this same idea to teleprocessing. 

Significance to Program of the Division: 

The most Important aspect of the NIH terminal system Is 
that an NIH standard communications language be defined. 
Once this Is done, any type of remote will be able to use 
the Computer Center's facilities. 

Such a system will give the Computer Center, DCRT, and 
NIH a pov/erful and flexible system. The capacities of 
the central system will be available to remote locations, 
all systems will be more reliable, and the computer load 
will be balanced dynamically rather than manually. 



41 



Proposed Course: 

1. Develop a plan for phased implementation of the 
Lavender Box concept in the Computer Center. The 
approach must be delicate: the current reliability of 
the Center's teleprocessing activities may not be 
jeopardi zed. 

2. A prototype of a SIN (system interchange module) is 
being built on the PDP-10 system. Experience with this 
implementation should be useful In subsequent development 
of the Lavender Box ideas. 

Honors and Awards: None 

Publications: None -^: — 



42 



Serial No. 2^J 

1. Computer Center Branch 

2. 

3. BethesHa 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 50, 1972 

Project Title: INTERFACE (technical communications) 

Previous Serial Number: Same 

Principal Investigator: Joseph D. Naughton 

Other Investigators: Staff of Systems Team and PAL Unit 

Cooperating Units: None 

Man Years: 

Total: 2 
Profess I onal : 1 
Others: 1 

Project Description: 

Objectives : 

1. The primary purpose of this effort is to give the 
Computer Center a regular means of conveying, (1) highly 
current technical Information to the widely dispersed 
computer programmers and systems analysts at NIH, and (2) 
current Center accomplishments, plans, policies, course 
offerings and other information of Interest to personnel 
and general managers at NIH. 

2. A secondary purpose is to foster communication and 
collaboration among computer users and between them and 
the Center by providing a forum for: (1) announcing key 
new projects, accomplishments, personnel or 
organizations; and (2) airing viewpoints or suggested 
approaches to computing problems. 

Methods Employed: 

INTERFACE is published every four to six weeks, and 

distributed to all scientific and administrative personnel 

who have expressed a desire to be kept up-to-date on 
computing at NIH. 

43 



Significance to Program of the Division: 

INTERFACE complements the Users Guide and other technical 
manuals by highlighting items and directing readers to the 
other publications for details. INTERFACE has given the 
users of the Computer Center a single reference point for 
all communications concerning the use of computers at NIH. 
It has done an excellent job of keeping the users Informed 
of new services, major systems changes and all facets of 
computing at NIH. 

Proposed Course: 

INTERFACE will continue to inform the Computer Center 
users of all pertinent computing highlights on a regular 
basis. 

Honors and Awards: None '■'- 

Publications: None 



44 



Serial No. 2jJ 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - N!H 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: IBM 2250 Graphic Program Support 

Previous Serial Mumber: Same 

Principal Investigator: James Oberthaler 

Other Investigators: Robert H. Brunelle 

Cooperating Units: None 

Man Months: 

Total: 1/2 
Professional: 1/2 
Others: 

Project Description: 

Object i ves : 

1. To maintain the NiH-produced PLl Graohi c Subrout ine 
Package which enables users to communicate with the IBM 
2250 Graphic Display using the PLl programming language. 

2. To instruct potential users In the Graphics 
capabilities offered by the Computer Center. 

3. To act as a consultant for programmers Implementing 
systems for the 2250 Display. 

Methods Employed: 

The PL/1 Graphic Subroutine Package is designed to enable 
the programmer to communicate with the IBM 2250-1 using 
OS PL/1, version 5. The programmer using this package 
can have access to the 2250-1, as well as the resources 
of PL/1. The ability to issue calls to procedures writ- 
ten in Assembly Language enables the programmer to send 



45 



data to and from the display. Interrupts are handled by 
user-defined PL/1 procedures. The subroutines included 
in the IBM System/360 Operating System Graphic Program- 
ming Services for the IBM 2250 Display Unit, Form 
C27-6909, are also made accessible to the programmer. 

Major Findings: 

1. The teaching of Fall and Spring semester courses in: 

Basi c Graphi cs , and, 
PLl Graphi cs . 

2. The Graphics Additions to Fortran (GRAF) package, 
developed at UCLA, was installed at the request of 
several users of CCB facilities. No programming 
support is available for this package, and so far no 
use of the system has been reported. 

3. The 2250 display unit was connected to the IBM 370/165 
vyhere response time for lengthy requests has been reduced 
by a factor of 60. 

Significance to Program of the Division: 

NIH users will be able to extend the data processing 
pov;er of System/360 computers: (1) to handle the graphic 
information associated with medical research and analysis 
applications; and, (2) to provide faster and more 
effective retrieval and graphic expression of medical 
data. 

Proposed Course: 

To continue support at the current level. 

Honors and Awards: None 

Pub! i cat i ons : None 



46 



Serial No. 2. 10 

1. Computer Center Branch 

2. 

5. Bethesda 



PHS - NIH 

Individual Project Report 

July I, 1971 through June 30, 1972 



Project Title: CPS - Conversational Programming System 

Previous Serial Number: Same 

Principal Investigator: John Camp 

Other Investigators: Trudy Kenny 

Cooperating Units: None 

Man Years: 

Total: 1-1/2 
Professional : 1-1/2 
Others: 



Project Description: 
ObjectI ves : 

CPS continues to provide much of the pov/er of the central 
computer facility to the researcher in his lab or office. 
By using CPS the researcher has available the computing 
capability of the central 360/370 computers as well as the 
ability to access data bases stored on-line at the 
central site. 

Methods Employed: 

The CPS system allows multiple users to simultaneously 
write, debug, and execute 360 computer programs conver- 
sationally. Programs are written in a subset of the PL/1 
language from a 27itl typewriter terminal or teletype 
located in or near the user's office. The terminals are 
connected with the NIH central 360/370 system over 
telephone lines on a dial-up basis. In addition, CPS 
supports a dialect of the BASIC language and provides a 
remote job entry (RJE) facility through which jobs can be 
submitted to be run In the background batch environment. 



47 



Major FI ndi ngs : 

Continuing enhancements to CPS place increasing power at 
the user's disposal. During the past year two 
maintenance releases of CPS have been installed. Each of 
these has increased the reliability of the system and 
added additional power to the language. Changes to the 
statement executor corrected certain ambiguities in the 
way statements v/ere executed. Modifications to the 
symbol processor effectively increased the number of 
symbols available within a program. Memory page handling 
v7as changed so as to minimize the build-up of page queues 
v/hen there are heavy loads on the system. 

NIH modifications to this IBM distributed program have 
enabled it to operate effectively as an integral part of 
the Computer Center. These modifications provide inter- 
faces for accounting, remote job entry, and dataset 
nami ng. 

CPS nov/ uses a special shared catalog for its dataset 
references so that batch jobs creating datasets for CPS 
can run on any machine. It is now possible to define 
groups of users who are allowed to allocate datasets on 
certain disk packs and to prevent other users from 
allocating on those packs. Extensions have been made to 
the ASCII character set available to teletype compatible 
termi nal s . 

During the past year there has been continuing growth in 
the number of users and in the use of CPS. Moving CPS 
from a 360/50 to a 370/165 greatly increased the potential 
capacity of the system and improved its responsiveness. 
The average number of concurrent users has approximately 
doubled during the past year. 

Significance to Program of the Division: 

CPS continues to provide computing capacity in the lab 
or office allowing the researcher to utilize the central 
computer facility without leaving his normal working 
environment. Thus the computer is more readily available 
and much time Is saved by not making the user come to the 
central site. By providing a truly Interactive conver- 
sational environment CPS gives the user capabilities 
which could never be achieved In a strict batch mode of 
operation. 

Proposed Course: 

Intended support includes continuing to provide mainten- 
ance for CPS. New versions and updates will be installed 
as they become available and are needed to Improve the 

48 



reliability and performance of the system. Local plans 
include steps to further integrate CPS into the operation 
of the central facility, improvement in user interaction 
with CPS/ and improved performance. 

With the increasing load on the system increasing demands 
are placed on system resources, particularly memory. 
Concurrently an investigation is in progress to evaluate 
the feasibility of running CPS as a processor under TSO 
to alleviate this problem and to provide a better base 
for the system. 

Honors and Awards: None 

Publ i cat i ons : None 



49 



Serial No. 2.11 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: PDP-10 Systems 

Previous Serial Number: Same 

Principal Investigators: R.F. Sproull/ G. Ricart 

Other Investigators: H. Vreenegoor, R. Grunby 

Cooperating Units: None 

Man Years: 

Total: 1.5 
Professional : 1. 5 
Others: 

Project Description: 

Obj ect 1 ves : 

Development of techniques and methods to make the PDP-10 
computer better suited to handle the computational needs 
of users for highly flexible, interactive systems. Make 
these services available on a f ee-f or-servi ce basis. 

Methods Employed: 

1. Continuing development and maintenance of the PDP-10 
monitor supplied by Digital Equipment Corporation. This 
time-sharing monitor has required several augmentations 
In order to meet NIH needs. Further, the manufacturer 
periodically updates this system, which requires that the 
new system be installed, together with NIH modifications. 

2. Continuing development and support of a library of 
systems programs to aid users. For example, the SOS text 
editor has been augmented to meet several new requirements 



50 



3. Continuing and renewed support of Interactive 
graphics capabilities using the PDP-10. The service 
programs for the 31+0 Display and RAND tablet have been 
redone In order to support more highly Interactive 
systems with less load on the timesharing system. Only 
the portion of a user program which actually drives the 
display (the "display file") Is retained In core; the 
balance may be swapped to secondary storage If other 
timesharing users require more Immediate service. 

k. Adding communications support to allow remote computer 
and terminal tie-lns for symbiotic usage of the PDP-10. 
These additions have required procurement of hardware for 
high-speed communications. Software Is being built to 
Implement Lavender Box message switching (See Serial 
Mo. 2.7). A high-speed link from the PDP-10 to the 
AGT-30 Graphics terminal has also been constructed. 

5. Encouragement of users to develop specialized packages 
to suit a particular class of needs. Sometimes these 
programs gain wide usage, as have MLAB (A Modeling and 
Graphics Package), and R. Feldmann's Structure Search 
Package. 

Major Findings: 

1. It is quite feasible to attempt systems modifications 
with the DEC system. 

2. Ways ought to be investigated to better coordinate 
NIH software changes with software updates from DEC. 

3. As the PDP-10 shifts from laboratory to fee-for- 
service environment, demands on systems staff for 
production programs and user support increase dramati- 
cally. Current staff is sufficient only to keep even 

vn th the essential workload and most important projects. 

Significance to Program of the Division: 

The PDP-10 has supported various DCRT programs with 
time-sharing services as v;el 1 as unique programming 
languages and graphics support. In addition, various 
smaller laboratory computers supported by DCRT on the 
PDP-10 for program assembly and computational support. 

Proposed Course: 

Continuing maintenance and support. Systems development 
as staff time permits. Upgrade and maintain software to 
keep pace v/ith user's needs. 



51 



Plans will be made for expansion of the PDP-10 as usage 
grows. The current KA-10 processor can be upgraded to 
DEC KI-lO's to provide for more computing power; similar 
growth is expected in available peripherals. 

Improvement of communications by increasing asynchronous 
line capacity and by acquiring synchronous communications 
equi pment. 

Graphics support will be expanded to drive a variety of 
display devices with one basic set of system subroutines. 
The DEC3U0 display^ the AGT30 display, and the COMPUTEC 
kOO display will constitute the initial set. 

Honors and Awards: None 

Publ i cat I ons : None 



52 



Serial No. 2.12 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Computer Training Program 

Previous Serial Number: Same 

Principal Investigator: Rita G. Minker 

Other Investigators: Patricia L. Logan 

Cooperating Units: All DCRT Laboratories and Branches 

Man Years: 

Total: 6.75 
Profess lonal : 6 
Others: 0.75 

Project Description: 

ObjectI ves : 

1. To provide the MIH staff with an understanding of and 
ability to use modern computer technology. 

2. To facilitate the formulation and analysis of the 
problems to be solved so that the technology can be 
useful 1 y appl I ed. 

Methods Employed: 

1. Training courses are tailored to the specific 
hardware and software available at NIH. 

2. Seminars focus on both the underlying disciplines 
(e.g., applied mathematics, computer-related engineering, 
information sciences) and on the specific areas of 
potential application. 

Significance to Program of the Division: 

Training courses enable Investigators to use DCRT's 
powerful equipment. Effective use of advanced techniques 

53 



and advanced equipment is obtained by further training 
and directed discussion. 

During tlie past 12 montliS/ over 2000 registrations were 
processed for the approximately 80 short courses offered. 

Proposed Course: 

Fall 1972 and Spring 1973 semesters of training classes 
and seminars are planned. 

"Introduction to CPS Files - A Self Study Guide" is being 
wr i tten. 

Honors and Awards: None 

Publ i cat ions : 

1. Computer Center Branch: Computer Training 
Courses--Fal 1 Term 1971. U. S. Dept. of Health, 
Education and Welfare, Public Health Service, National 
Institutes of Health, Division of Computer Research and 
Technology, Computer Center Branch. 

2. Computer Center Branch: Computer Training Courses 
and Semi nars--Spr i ng Term 1972. U. S. Dept. of Health, 
Education and IVelfare, Public Health Service, National 
Institutes of Health, Division of Computer Research and 
Technology, Computer Center Branch. 

3. Computer Center Branch: CPS PL/I Reference 



54 



Serial No. 2.13 

1. Computer Center Branch 

2. 

3, Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: On-line File Processing System 

Previous Serial Number: Same 

Principal Investigator: Jon A. Halverson 

Other Investigators: Marvin Katz 

Cooperating Units: None 

Man Years: 

Total: 2 
Professional : 2 
Others: 

Project Description: 

Obj ecti ves: 

The primary objective Is to establish an on-line file 
processing capability at the NIH whereby the user can 
maintain, retrieve and display information in a data 
base via a remote terminal. 

Methods Employed: 

The system being Implemented is a modularly constructed 
program designed to provide the basic control system 
program for the Installation of on-line information 
systems. It functions as an Interface between user- 
written processing programs and the IBM system/360 
operating system. 

Major Findings: 

The monitor to provide the data base/data communications 
interface has been Implemented. To evaluate its effec- 
tiveness, the activities of the NIH Clinical Center 
Admissions Department has been chosen as the first 
application. This subsystem is now operational. 

55 



Significance to Program of the Division: 

The system can be considered as an initial step toward 
the development of on-line information systems for the 
MIH. For the first time investigators responsible for 
maintaining large data bases v;lll have a pov/erful means 
of accessing critical Information quickly. 

Proposed Course: 

Attention will be given to operating the system in an on- 
line, multi-user environment. Special emphasis will be 
given to applying the system to other areas In the NIH 
Clinical Center, In addition to Admissions, which can 
benefit from interactive processing. 

Honors and Av/ards: None 

Publications: None 



56 



Serial nq. 

1. Computer 



2. 
3. 



Bethesda 



i.m 



Center Branch 



PHS - NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 



Project Title: Optical Document Processing System 

Previous Serial Number: Same 

Principal Investigator: Elliott Alterman 

Other Investigators: John Camp 

Cooperating Units: None 

flan Years: 



Total : 

Professional 

Other: 



0.1 
0.1 




Project Description: 
Objectives : 



NIH processes vast volumes of data in many forms and 
formats. The effort necessary to analyze and process the 
data currently requires significant time and effort to 
convert these data to machine readable form. We have 
eliminated much of this intermediate data conversion by 
using the documents themselves as input to the central 
facility computers. 

Methods Employed: 

Internal processing routines have been developed to pro- 
cess documents read by the optical character recognition 
equipment (IBM 1288 Optical Page Reader). Each document 
Is described uniquely in terms of a set of parameters. 
This information Is maintained in an on-line data set 
accessible to the processing program. The flexibility of 
this design is displayed In the elminatlon of the 
requirement for writing new processing programs for each 
new document we wish to process. Changes to document 
design, processing options, or output formatting can be 
done in a matter of minutes. 

57 



Major Fi ndi ngs : 

This system permits rapid and efficient collection of 
data for processing by a computer. Significant cost 
reductions have been achieved in the year that the system 
has been functioning. Coupled with the great increase in 
data availability this has yielded a high performance, 
low cost data collection tool that permits greater 
utilization of available data. 

Significance to Program of the Division: 

The Division of Computer Research and Technology has a 

major goal of making access to the central computer 

facility easy, simple and cost effective. The Optical 

Page Reader System provides an avenue for that access 

in a manner requiring minimal change to external operating 

procedures. 

Honors and Awards: None 

Publications: None 



58 



Serial No. 2.15 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Chemical Structure Search and Display 

Previous Serial Number: None 

Principal Investigator: R.J. Feldman 

Other Investigators: S.R. Heller 

Cooperating Units: PRA-NCI, LM-NHI 

Man Years: 

Total : 0.6 
Profess ional : 0.6 
Other: 

Project Description: 

Objecti ves : 

The improvement of techniques for storage, search and 
display of chemical structures encoded as connection 
tables. The Implementation of various aspects of a user 
oriented chemical information system. 

Methods Employed: 

Programs already in existence were brought to completion. 
New programs were written and debugged as our understand- 
ing of the problem changed. 

Major Findings: 

A new method for structuring chemical structure files was 
developed resulting in the ability to do interactive 
chemical sub-structure search for large files. 



59 



Significance to Program of the Division: 

The chemical structure search program Is now being tested 
for use by the cancer chemotherapy program in the NCI and 
by the mass spectrometry lab in the ^!HLI . 

Proposed Course: 

Improvements to the search program depend heavily on use 
by scientists. Recommendations for improvements v/ill be 
Incorporated into the program. 

I'Jork will be done to establish structure files of 
Interest to different groups of scientists. In 
particular, preparations will be made for establishing 
the complete chemical abstracts (CAS) registry file of 2 
million compounds. The existence of the complete 
registry file will guarantee a scientist doing a chemical 
structure search that the search results are meaningful 
and complete. 

Honors and Awards: None 

Publ i cations : 

Feldmann, R.J.; Heller, S.R. - An Application of 
interactive Graphics - The Nested Retrieval of Chemical 
Structures, Journal of Chemical Documentation, Vol. 12, 
No. 1, 1972, pp U8-5tt. 

Feldmann, R.J.; Heller, S.R.; Shapiro, K.P. - An 
Application of Interactive Computing - A Chemical 
Information System, Journal of Chemical Documentation; 
Vol. 12, No. 1, 1972, pp Ul-47. 



60 



Serial No. 2.16 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Macro Molecular Display and Manipulation 

Previous Serial Number: None 

Principal Investigator: R. J. Feldmann 

Other Investigators: J.S. Cohen, C.R.T. Bacon, S.R. Heller 

Cooperating Units: LMB-MIAMD 

Man Years: 

Total 0.6 
Professional : 0.6 
Other: 

Project Description: 

Objecti ves : 

The development of a macro molecular display system using 
the Adage AGT-30 Graphics terminal and the PDP-10 
computer. 

Methods Employed: 

Programs for molecule file generation, editing and 
display were written and debugged. Consideration in 
designing the system was given to the needs of the users. 

Significance to Program of the Division: 

Scientists can now view and manipulate macro molecular 
models. Computer drawn diagrams of molecule sections for 
inclusion in publications can now be easily and quickly 
generated. 

Honors and Awards: None 

Publications: None 



61 



Serial No. 2.17 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Dial-Up Graphics 

Previous Serial Number: Same 

Principal Investigator: R. J. Feldmann 

Other Investigators: None 

Cooperating Units: None 

Man Years: 

Total: 0.1 
Professional : . 1 
Others: 

Project Description: 

Objecti ves : 

The development of a computer terminal with graphics 
capability. Communication with the computer (PDP-10) to 
be over dial-up telephone lines. 

Methods Employed: 

The characteristics of existing computer terminals are 
evaluated for performance with respect to the needs of 
projects within the division. 

Major Fi ndi ngs : 

The state of technology is developing more slowly than 
v/as ant i ci capted. The silicon target storage tube is not 
acceptable for use in a graphics terminal. 

Proposed Course: 

There is still a need for a remote dial-up graphics 
terminal. Continuing evaluation of new and improved 



62 



graphics units will be done In an attempt to find an 
acceptable unit. 

Honors and Awards: None 

Publ icat ions : None 



63 



Serial No. 2.18 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - MIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Computer Output Microfilm 

Previous Serial Number: None 

Principal Investigator: William Speary ~ 

Other Investigators: F. Rosenzweig/ W. Jones, R. Baxter 

Cooperating Units: None 

Man Years: 

Total: 1/2 

Profess ional : 1/2 
Others: 

Project Description: 

Obj ect i ves : 

To provide the NIH community a mechanism to obtain high 
volume output in a manner by v«;hich the output can be 
easier stored, manipulated and processed at a considerable 
cost and manpower saving. 

Methods Employed: 

A Datagraphix Micromation Recorder Model t+f+eo is being 
acquired by DCRT/CCB. General -pur pose software routines 
are. being written to allov.' the NIH community to obtain 
output which now Is produced on paper to be produced on 
16mm microfilm or 105mm microfiche. 

Significance to Program of the Division: 

Users will be able to obtain large volume computer output 
on a medium v/hlch Is: (1) easier to handle; (2) easier 
and cheaper to store; and (3) at a significant cost 
reduction compared to printed output. 



64 



Honors and Awards: None 
Publ i cations : None 



65 



Serial No. 2.19 

1. Computer Center Branch 

2. 

3. BethesHa 



PHS - MIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Installation of Time Sharing Option (TSO) of 
OS/360 MVT 

Previous Serial Number: None 

Principal Investigator: William Speary 

Other Investigators: A.C. Curtis , 

Cooperating Units: None 

Man Years: 

Total: 1/U 

Professional : 1/k 
Others: 

Project Description: 

Obj ect I ves : 

To provide a Time Sharing System based on central 
computing facility for the NIH community to enable them 
to implement data processing programs In a shorter period 
of time. 

Methods Employed: 

OS/360 MVT-TSO has been installed in the NIH central 
computing facility. It is currently undergoing tests and 
modifications in order to Integrate It Into the NIH 
envi ronment. 

Significance to Program of the Division: 

Laboratories will be able to have, via terminals, 
computing facilities which have heretofore been available 
only through the 'batch' computer service. TSO also 
allowed the laboratories to program/debug/ impl enent 
programs via terminals which are unique to the individual 
1 aborator I es . 

66 



Proposed Course: 

CCB Is in the process of modifying/enhancing TSO In 
order to operate in the NIH environment. 

Honors and Awards: None 

Publ I cat i ons : None 



67 



July 1, 1971 through June 30, 1972 

STITUTES OF HEALTH 
MD TECHNOLOGY 

Summary of Branch Activities 1. dCRT - 3 



PUBLIC HEALTH SERVICE - NATIONAL INSTITUTES OF HEALTH 
DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 



2. LABORATORY OF APPLIED STUDIES 3. Eugene K. Harris 

Chief 

I. GENERAL SU^MARY 

The Laboratory of Applied Studies engages primarily in cooperative studies with 
NIH scientists, applying mathematics, statistics, and computing science to bio- 
medical problems. To support these studies, special -purpose computer pro- 
gramming systems are developed where needed; these are often later extended for 
more general use and distributed to potential users in the form of DCRT 
technical reports. In addition, the Laboratory undertakes mathematical and 
mathematical statistical research with applications to biological studies. 

Continuing last year's format, each project report includes a number of "sub- 
projects" contributing to the same general objective. 

II. HIGHLIGHTS OF CURRENT LABORATORY PROGRAMS 

1. Clinical Projects 

This past year has seen a continuation and strengthening of collaborative 
research in three areas: a) evaluation of computer programs for the diagnosis 
of electrocardiograms, b) development and application of computer-assisted 
methods for continuous analysis of ventricular volume and pressure in patients 
with ventricular disorders studied by angiography, and c) computer and mathe- 
matical support of diagnostic and research studies of the Nuclear f^dicine 
Department of the Clinical Center. 

In cooperation with the Cardiology Branch, NHLI, over 1,000 selected ECG's 
have been analyzed manually and through two computer programs: the Nb-yo Clinic 
and Public Health Service programs. Use of a new IBM program for ECG diagnosis 
began this year. Statistical evaluation of results, supported by catheteriza- 
tion data where available, is in progress. The computerized system for digital 
analysis of analog cineangiographic data, developed last year to study dynamic 
changes in ventricular volume and related parameters, is now being used in the 
care of patients with coronary artery disease. In particular, algorithms have 
been written to assist in analysis of left ventricular wall motion in patients 
before and after bypass surgery. 

A new, small computer system to receive and store scintillations from the 
Gamma cameras of the Nuclear Medicine Department has been procured and is now 
undergoing acceptance and performance tests. During this year, LAS has 



69 



cooperated with the Nuclear Medicine Department in the development o£ programs 
to evaluate cerebral and hepatic blood flow. 

Various statistical studies in the area o£ clinical chemistry have been 
completed or progressed during this year. These have been independently 
pursued although the database derives from earlier collaborative work with the 
Clinical Chemistry Service, C. C. Studies completed and published include 
analysis of normal variations in ionized calcium, and development of a general 
mathematical theory to explain the distributions of many clinical measurements 
from which normal ranges are derived. Analysis of patient data, in coopera- 
tion with various I/D Laboratories, has begun, but progress is slow, largely 
because of the need at this stage to explore the usefulness of a variety of 
statistical methods not usually applied to clinical research - e.g., time 
series analyses, smoothing and quality control methods. 

2 . Laboratory Research Projects 

In the last Annual Report, a significant advance was noted in establishing 
the mathematical equivalence of the Scatchard model of macromolecule binding to 
the step-wise equilibrium equations. This has enabled the reanalysis of much 
published data to estimate dissociation constants and determine classes of 
binding sites in macromolecule -ligand reactions of physiologic importance. Such 
investigations were carried out this year on the binding of a long-chain fatty 
acid to serum albumin and a medium-chain acid to erythrocytes. The latter study 
in particular, has led to a better understanding of the mechanisms v^ereby free 
fatty acids are bound by mammalian cells. 

A general model of capillary blood flow and tissue diffusion, accounting 
for the possible effects of periodic flow on tissue concentrations, has been 
completed, and numerical solutions obtained. 

Two neurophysiologic projects have made progress this year. One involves 
experimental verification and computer modelling of the time relationships 
between cerebellar input and firing of Purkinje cells (cerebellar cortex) and 
sub -cortical nuclear cells. The present year has been largely concerned with 
data collection through experiments performed in cooperation with the 
Laboratory of Neurophysiology, NIIVH. The sub-cortical nuclei have been found 
to be directly responsive to mossy fiber input, rather than requiring 
mediation by Purkinje cells. In the second study, a hardware -simulated neural 
model, capable of reproducing the characteristics of a small neural net with 
distributed dendritic input, has been incorporated into the LAS computer 
(MAC-16) . Input to the model, monitoring, analysis and display of action 
potentials, spike trains, etc., can now be entirely computer -controlled. 
Current plans call for applying this powerful combination to the study of small 
neural nets in cooperation with the Laboratory of Neural Control, NINDS. 

3. Research in Mathematics and Statistics 

The Biomathematics and Statistics Section of the Laboratory has continued 
its work on the structure of different multivariate distributions involved in 
the selection of variables for study of size and shape. The theory has been 



70 



applied to an evaluation of discriminant analysis in a genetic study of human 
skull proportions, in cooperation with NIDR. As a by-product of this research, 
computer programs have been prepared and are now available for testing the 
statistical power of the multiple correlation coefficient, and for related 
confutations in multivariate analysis. 

4. Education, Training, and Awards 

During the past year, Laboratory staff conducted courses for NIH scientists 
in the areas of graph theory, data structures for modelling, use of the Calcomp 
plotters and the design of assemblers and loaders. 

J. IVbsiinann, Head, Biomathematics and Statistics Section, chaired a sub- 
committee on training for mathematicians and statisticians of the HEW Career 
Service Board for Mathematicians and Statisticians headed by Dr. T. Woolsey, 
Director, NICHS. A report on training in HEW was completed. 

Dr. Mosimann was named a Fellow of the American Statistical Association 
in August, 1971 . 



71 



LAS reports published or "in press" during FY 72 



1. Harris, E. K. , and DeMets, D. L.: Biological and analytic components o£ 
variation in long-term studies o£ serum constituents in normal subjects, 

V. Estimated biological variations in ionized calcium. Clinical Chemistry . 
17: 983, 1971. 

2. Harris, E. K., and DeMets, D. L. : Effects of intra- and inter -individual 
variation on distributions of single measurements. Clinical Chemistry . 
18: 244, 1972. 

3. *Hutchinson, G.: Modular lattices and abelian categories. J. of Algebra . 

19: 156, 1971. 

4. Larson, S. M. , Millar, R. C, Chalmers, T. C, Bailey, J. J. , and 
Johnston, G. S.: Hormonal influences on hepatic and portal vein perfusion 
beds. Gastroenterology . April, 1972. 

5. *Magar, M. E., Steiner, R. F., and Fletcher, J. E. : Analysis o£ protein 

ligand equilibria. J. Theor. Biol . 32: 59, 1971. 

6. Marcus, M. L., Schuette, W. H., Whitehouse, W. C, Bailey, J. J ., and 
Clancy, D. L. : An automated method for the measurement of ventricular 
volume. Circulation . 45: 65, 1972. 

7. Mprtimer, J. A ., and Evarts, E. V. : Latency differences in cerebellar | 
pur kin je and nuclear cell activity in association vdth startle responses. 
Federation Proceedings . 31: 378, 1972. 

8. Specter, A. A., Fletcher , J . E . , and Ashbrook, J. D . : Analysis of long- 
chain free fatty acid binding to bovine serijm albumin by determination 
of stepmse equilibrium constants. Biochemistry . 10: 3229, 1971. 

9. Specter, A. A., Ashbrook , J . D . . Santos, E. C, and Fletcher, J. E . : 
Quantitative analysis of free fatty acid uptake by mammalian cells: 
Laurie acid and human erythrocytes. J. of Lipid Research . 1972. 
(in press) . 

10. Woody, C. D. , Harris , E. K. , and Brozek, G. : Background noise content of 
evoked responses from facial nucleus of cat. Intern. J. Neuroscience . 
1: 361, 1971. 



cited in last year's annual report as being "in press" 



72 



Serial No. DCRT 3.1 

1. Laboratory of Applied Studies 

2. Biomathematics and Statistics 

Section 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Biomathematics and Statistics 

Previous Serial Number: Same 

Principal Investigator: J. E. Mosimann, G. Hutchinson 

Assisted by: G. Atta 

Co- Investigators: J. Niswander, NIDR 
A. Coulombre, NEI 

Man Years: Professional - 3.0 
Other - 0.2 

Ob j ectives : 

To conduct mathematical statistical research in areas of biological 
importance, especially in multivariate problems, and to develop related 
computing methodology of use to biostatisticians and other health 
scientists. To investigate the applicability of mathematical theory, 
particularly finite mathematics including graph theory to various bio- 
medical problems . 

Progress During Past Year: 

Theoretical work on the development of multivariate mathematical methods 
for size and shape analysis has led to an understanding of covariance 
structure in various families of mathematical distributions. Thus, 
certain families of distributions contain no members for which shape is 
independent of certain size variables. For example, no member of the 
multivariate lognormal family can have shape independent of the sum of 
the variables . 

Shape methods have proved to be useful adjuncts to discriminant analysis 
in the study of human skull proportions (NIDR) . They made it possible 
to avoid errors in discriminant analysis resulting from too many vari- 
ables relative to the number of observations. 

Various statistical methods were applied to the analysis of growth of 
scleral ossicles in the eyes of chick embryos (NEI) . 



73 



Studies of statistical power for tests of independence of size and shape 
using the multiple correlation coefficient were undertaken. Programs 
prepared in connection with this project comprise area-to-point and 
point-to-area programs for the Normal, Chi-squared, F and Non-central 
F distributions. Supplementary programs evaluate Pearson's Incomplete 
Beta Function ratio and the unconditional power of the multiple 
correlation test. 

The procedures for the normal, chi-squared, F , and non-central F 
distributions compute the area under the respective curves from -°° to 
any given point. The area-to-point programs compute the point marking 
off any given area from -«> to that point. 

A program for converting time -sharing computer terminal input into 
printed mathematical text output using the MN print train was prepared. 
Documentation has been prepared and is continuing. 

Using linear programming and graph theory, a study of chemical reaction 
mechanisms was completed. In particular the collapse of reactions due 
to the exhaustion of substances in the system has been studied. Results 
indicate the inadequacy of complex "pathway" models of biochemical 
systems which fail to take into account this possibility of terminating 
reactions . 

Proposed Course: 

1. Studies of dispersion structure in non-normal multivariate distri- 
butions; applicability to non-normal biomedical data. 

2. Continuation of studies of the power of multiple correlation tests. 

3. Development of Covariance -Correlation and other ratio and proportion 
programs for data analyses (NIDR, NEI) , including discriminant 
analysis. 

4. Studies of linear algebra and graph theory with programming 
techniques for symbol manipulation; applications to mathematical 
text editing and medical records. 

Publications: 

1. Hutchinson, G. : Modular lattices and abelian categories. 
J. of Algebra. 19: 156-184, 1971. 



74 



Serial No. DCRT 3.2 

1. Laboratory of Applied Studies 

2. Office of the Chief 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Evaluation of Computer-assisted Systems for Patient Care 

Previous Serial Number: Same 

Principal Investigators: J. J. Bailey, M. Horton 

Assisted by: M. Douglas, M. Gelfand 

Co -Investigators: S. Epstein, S. Itscoitz, L. Grauer, J. Hirshfeld, R. Karsh 
CS, NHLI 

G. Johnson, E. Jones, M. Green, S. Larson, NM, CC 
D. Syed, H. Ostrow, CSL, DCRT 
T. Jones, HL, DCRT 
J. Prewitt, DMB, DCRT 

Man Years: Professional - 2.0 
Other - 0.3 

Project Descriptions: 

Two sub-projects are currently being conducted in the general area of the 
evaluation of health care systems, particularly those concerned with cardio- 
vascular problems. These projects are: a) evaluation of computer diagnosis 
of electrocardiograms (ECG) ; b) computer-assisted studies in nuclear 
medicine. These projects were introduced in last year's report. 

a) Automated ECG Analysis 

Objectives: 

To assess on a sound statistical and clinical basis the usefulness of 
computer programs for the diagnosis of electrocardiograms. ECG's from 
a range of cardiovascular disease, as well as presumed normal, are 
selected in cooperation with the Cardiology Branch, NHLI. 

Progress during FY 72: 

Analog-to-digital conversion which in FY 1971 was implemented on the 
IBM1827 with an unacceptable error rate, has been successfully reimple- 
mented on the MAC-16 system without significant error. Over 1000 
electrocardiograms have been digitized with this system and processed 
by Mayo Clinic and Public Health service programs. Processing by a 
new IBM program has been arranged. A controlled statistical evaluation 



75 



and comparison o£ the programs is now being prepared for publication. 

Proposed Course: 

Implementation of a program developed by the VA Research Center for 
Cardiovascular Data Processing will be complete in FY 72 and 
evaluation in FY 73. Selected electrograms in digital form from 
patients in a five year hyperlipidemia study (Dr. J. Brensike, Lipid 
Branch, NHLI) are routinely set aside for further analysis by which- 
ever of the four programs above is deemed appropriate. 

b) Computer Assisted Studies in Nuclear f^dicine 

Objectives: 

To provide computer-based mathematical analysis in support of 
diagnostic and research activities of the Nuclear Medicine Depart- 
ment, C. C. 

Progress during FY 72: 

After a year of reviewing computer applications to nuclear medicine, 
and several site visits, a joint lAS-CSL team drew up stringent 
specifications for a small computer to be dedicated to the gamma 
cameras in the Nuclear Medicine Department of the Clinical Center. 
This team then evaluated several proposals and selected a vendor, 
who built the specified system under monitoring by the joint LAS- 
CSL team. The system, delivered in April, 1972, is undergoing 
extensive software and hardware checkout by a joint DCRT -Nuclear 
Medicine task force. 

, , A program to estimate regional blood flow in an organ from the count 
density of the organ and from its arterial input has been developed 
and applied to cerebral blood flow studies (ref . 1) . A method to 
estimate radionuclide washout rates in monkey liver preparations was 
developed using the MLAB package on the PDP-10 computer. This 
computer processing allowed researchers to investigate hepatic blood 
flow under the influences of various hormones (refs. 2 and 3). 

Proposed Course: 

The computer will be used to map expected densities, based on 3- 
dimensional organ models, to continue blood flow studies, and to 
explore mathematical methods of image improvement. 



76 



Reports : 

1. Larson, S. M., Bailey, J. J., Griffith, J. B., Shall, G. L., and 
Johnston, G. S.: Radio-isotope techniques for measuring regional 
blood flow. Submitted to Internet '1 J. Appl. Radiation and Isotopes . 

2. Larson, S. M. , Millar, R. C., Chalmers, T. C., Bailey, J. J., and 
Johnston, G. S. : Hormonal influences on hepatic and portal vein 
perfusion beds. Gastroenterology , April, 1972. 

3. Millar, R. C., Larson, S. M. , Bailey, J. J., Johnston, G. S., 
Chalmers, T. C, and Ketcham, A.: Estimation of hepatic blood flow 
using Xenon 133: A chronic surgical preparation. Submitted to 
Surgical Res. 



77 



Serial No. DCRT 3.3 

1. Laboratory of Applied Studies 

2 . Office of the Chief 

3. Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Computer Methods in Cardiologic Research 

Previous Serial Number: Same 

Principal Investigators: J. Bailey, M. Stroot, M. Gelfand 

Assisted by: M. Douglas, W. Scott 

Co -Investigators: D. Clancy, S. Epstein, CB, NHLI 

W. Whitehouse, W. Schuette, CC, BEI 

J. Mason, NASA 

H. Blum, DCRT 

Man Years: Professional - .75 
Other - .5 

Project Description: 

A. Automated Angiographic Ventricular Volumes and Ventricular Wall Motion 

Objectives: 

To develop and apply computer-based methods for determining the 
dynamic cardiovascular status of patients. 

Progress during FY 72: 

The LAS contribution to this study of ventricular angiography 
began in FY 71 with the application of computer operations to 
the recovery of continuous digital volume and pressure informa- 
tion from analog cineangiographic data on magnetic tape. 

Description of the data recovery, noise suppression and display 
systems were given in last year's annual report. 

Pressure and geometric data on 13 patients developed by this 
system have been used for calculation of myocardial wall stress, 
Csee part C) . 

During this reporting year, development of algorithms is contin- 
uing for the computer-assisted analysis of regional disorders 
of left ventricular dysfunction as shown by ventricular angio- 
graphy in patients with coronary artery disease. Efforts have 



78 



been made in the past to manually analyze disorders of left ventricular 
wall motion such as aneurism, akinesia, hypokinesia, dyskinesia, and 
asyneresis. This study analyzes these abnormalities in a more automated 
and impartial fashion, focussing on those patients who have been studied 
with ventricular angiography before and after coronary artery saphenous 
vein bypass graft surgery. 

Left ventricular angiograms are projected at 30 frames/ second and their 
outlines drawn by hand for one or two cardiac cycles. These drawings are 
converted to paired analog signals each representing one-half of the 
ventricular profile on a specially constructed video scanning device in 
the NIH Television Laboratory. An analog signal representing volume 
calculated on the basis of Simpson's Rule is also recorded. These 
signals are digitized on the CDC 3100 Hybrid Computer and all calculations 
of volume, area, length, and diameter are made on the IBM 370 computer. 
Analysis of ventricular wall motion is carried out on the PDP-10 computer. 

Proposed Course: 

These data will be used in Project C: Studies in Myocardial Stress. 

Publication: Marcus, M. L., Schuette W. H., Whitehouse, W. C, Bailey, J. J., 
and Clancy, D. L.: An automated method for the measurement of 
ventricular volume. Circulation . 45: 65, January 1972. 

B. Studies in Myocardial Contractility 

Objectives: 

To test and compare the reliabilities of proposed indices of myocardial 
contractility. Such indices include isometric contraction time, 
systolic ejection time fraction, mean systolic pressure, maximum 
change in pressure (max dp/dt) during isometric contraction, maximum 
dp/dt divided by the pressure (p) at which it occurs and maximal 
velocity of the contractile element (Atoax;) , (e.g., the Sonnenblick 
or Mason indices) . From such parameters may be computed indices 
related to ventricular power or the efficiency of power conversion. 

Progress during FY 72: 

In the cardiac catheterization laboratory, patients regularly studied 
by the Cardiology Branch of NHLI have simultaneous recordings of 
electrocardiogram, left ventricular pressure, and peripheral artery 
pressure. These analog records are digitized on the CDC 3100 Hybrid 
Computer and the digitized output is analyzed to isolate the period 
between the occurrence of the end-diastolic pressure and opening of 
the aortic valve for each cycle. Contractile element velocity (Vce) 
is computed over this interval on the basis of pressure alone, and a 
straight line is fitted to the portion of the Vce curve between the 
maximum Vce for each cycle (max Vce) and the time of aortic valve 



79 



opening. This is projected to the point where ventricular pressure 
equals zero to yield Vmax ;\iiich is also calculated in a similar 
fashion over a period of 0.05 sec. ignoring aortic valve opening. 
The appropriate part of the Vce curve and the fitted straight line 
are plotted for each 'Wnax calculation on the CALCOMP plotter. Max 
Vce, maximum dp/dt, and dp/dt at a developed pressure of 40 mm. Hg 
are also calculated. Four cycles are analyzed in each patient in 
normal sinus rhythm and ten in the case of atrial fibrillation. 

These measurements can be done as a part of the routine cardiac cathe- 
terization procedure and require only the insertion of the catheter - 
tip manometer into the left ventricle. Each has been proposed as a 
measure of left -ventricular contractility and this study allows the 
comparison of each of them with one another and with other measure- 
ments of ventricular performance. The computation of these parameters 
has also been included in a study of aortic valve disease in progress 
at the Cardiology Branch, NHLI. Fifty (50) patients have been 
analyzed in this fashion. 

Proposed Course: 

The independent measurements of end-diastolic pressure, cardiac index, 
end-diastolic voltmie, and ejection fraction made on the same patients 
by Cardiology Branch personnel will be compared with the various 
measurements of contractility. The relationship of myocardial con- 
tractility to RR- interval in patients with atrial fibrillation will 
be studied. The stibsequent measurements of cardiac performance made 
at post -operative catheterization on those patients who have come to 
operation will also be related to the contractility measurements in 
an attempt to determine their usefulness in predicting the operative 
outcome of valve replacement operations. 

Studies in Myocardial Stress 

Objective: 

Another important index of myocardial contractility may be derived 
from stress-strain relationships in the wall. This study evaluates 
and conpares proposed stress formulas. 

Progress during FY 72: 

In FY 72 the literature on myocardial stress was reviewed. Ten 
separate stress formulations were evaluated from a theoretical 
(mathematical) viewpoint and were compared using the pressure and 
geometric data from the system described in Project A. However, 
this comparison could be accomplished only for mean equatorial stress. 
The general stress formulation requires a numerical solution to a 
complex system of differential equations. A general expression for 
the stress tensor at any point in the myocardial wall may not be 



( 



80 



possible by the usual mathematical techniques for solving differential 
equations. A way around this problem may be the finite element analysis 
program developed by NASA to determine stress in nose cones of missiles. 
During FY 72, a joint LAS-NASA team concluded that application of the 
finite element analysis method to myocardial stress is appropriate and 
feasible . 

Proposed Course: 

The NASA program (IBM 360-95) at Goddard Space Center will be adapted for 
this purpose. Evaluation of a general stress formula will proceed using 
the pressure and geometric data described in Project A. 



81 



Serial No. DCRT 3.4 

1. Laboratory o£ Applied Studies 

2 . Office of the Chief 

3. Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Statistical Research in Clinical Pathology 

Previous Serial Number: Same 

Principal Investigators: Eugene K. Harris, David DeMets 

Assisted by: Willie Mae Scott 

Co -Investigators: D. Young 

F. C. Bartter and staff, EB, NHLI 

Man Years: Professional - 1.5 
Other - 1.0 

Project Description: 

Background: 

The studies of variation in normal blood chemistries which form the 
background of this project have been amply discussed in earlier 
annual reports. 

Progress during FY 72: 

The study of biological variation in serum ionized calcium, noted 
as underway in the preceding Annual Report, has been completed ^d 
published (1) . Intra -individual physiological variation in Ca 
was shown to be on the borderline of detectability, even with the 
analytic precision expected of modem ion-selective electrode 
systems . 

A study of the underlying statistical basis for "normal ranges" 
commonly used in clinical practice was begun during this reporting 
year. IVbst distributions of observations from which such reference 
limits are derived are non-gaussian in shape: skewed and, usually, 
peaked. A conditional probability model, including general expres- 
sions for intra- and inter -individual variation has shown (2) that 
such non-gaussian characteristics arise from correlations between 
means and variances as well as general heterogeneity among intra- 
individual variances. Methods to restore gaussian form by reducing 
these factors are being developed and implemented in computer pro- 
grams. This work is expected to provide a theoretical basis for the 



82 



proper treatment of observations to estimate normal ranges and proportions 
in any clinical variable. 

Studies have been initiated in cooperation with the Endocrinology Branch, 
NHLI, to investigate time -dependencies in blood pressure measurements in' 
hypertensive patients undergoing therapy. Results so far have been 
suggestive of periodic influences obscured by a high level of noisy fluc- 
tuations, probably due to immediate responses to extraneous stimuli. 

Proposed Course: 

Studies on the optimal univariate statistical analysis of measurements 
from normal individuals will be completed. A study of covariation among 
physiologically related variables in normal volunteers will be initiated 
with the goal of developing and testing statistical criteria for diag- 
nostic indexes combining measurements of two or more clinical variables. 
Investigation of time -dependencies and other properties of repetitively 
measured clinical quantities in patients and normal volunteers will be 
pursued more widely in cooperation with laboratories in several 
Institutes. It is hoped that this work will encourage the use of more 
sensitive methods in clinical research to evaluate the short-term effects 
of therapy. 



Publications: 



Harris, E. K. , and DeMets, D. L.: Biological and analytic components 
of variation in long-term studies of serum constituents in normal 
subjects, V. Estimated biological variations in ionized calcium. 
Clinical Chemistry . 17: 983, 1971. 

Harris, E. K. and DeMets, D. L.: Effects of intra- and inter- 
individual variation on distributions of single measurements. 
Clinical Chemistry . 18: 244, 1972. 



83 



Serial No. DCKT 3.5 

1. Laboratory of Applied Studies 

2. Applied Mathematics 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Applied Mathematics Unit 

Previous Serial Number: Same 

Principal Investigators: J. Fletcher, J. Ashbrook, E. Hill 

Assisted by: K. Drew 

Co -Investigators: A. Specter, Univ. of Iowa (formerly with Lab. of Metabolism, 
NHLI, under D. Steinberg); B. Hubbard, Univ. of Maryland; 
R. Shrager, PSL, DCRT. 

Man Years: Professional - 3.0 
Other -0.5 

Project Descriptions: 

Background and Objectives: 

The primary responsibility of the Applied Mathematics Unit is to 
provide NIH scientists with mathematical competence for biomathe- 
matical modelling and data analysis. This competence includes 
both theoretical and applied techniques, as well as numerical com- 
putation methods. Each individual in the unit has a primary 
specialty in computer science or mathematics, and each is a capable 
computer programmer. 

Progress during FY 72 : ' . 

1) . Macromolecule-Ligand Binding: 

A study of the binding of long -chain fatty acid to bovine serum 
albumin has been completed and published. The binding of medium- 
chain fatty acid to human serum albumin has been analyzed in terms 
of multiple stepwise equilibria and is being prepared for publica- 
tion. 

An investigation of the interaction of human erythrocytes with a 
medium-chain fatty acid (lauric) has been completed and is in press. 
Our research produced a completely new picture of the utilization 
of free fatty acid by mammalian cells. We were able to detect a 
hitherto unrecognized class of binding sites and obtained association 
constants which were different from what was previously supposed. 



84 



2) . Modelling of Biological Diffusion Processes 

The model for capillary flow and diffusion of substrates in the micro- 
circulation has been refined and completed. After a survey of existing 
numerical techniques for solution of the equations revealed no appro- 
priate methods, new techniques were developed within the Unit These 
methods provide an unconditionally stable implicit scheme for solving 
the mathematical model. The numerical errors are currently being evalu- 
ated and simulation of biologically important cases has begun. 

3) . General Support Activities 

a. Modelaide: 

Extensive revisions have been made to the MODELAIDE programs which 
were origijially written by R. Shrager (Physical Sciences Lab.) 
Information on changes and additions as well as instructions for 
using the system have been distributed in the form of a User's Guide. 
A member of the Applied Mathematics Units has assumed responsibility 
for maintaining and updating the software and the documentation, 
including the graphics programs. He also assists R. Shrager in 
conducting a DCRT Training Course dealing with the curve -fitting 
system. 

b. Personnel of the Unit are directly engaged in intramural training of 
computer users. Such activities include the courses: Data Structures 
for Modelling, Design of Assemblers and Loaders, Introduction to the 
Calcomp Plotter. Staff members also participate as instructors during 
their non-duty hours in the Upward Mobility Program for Advanced 
Education. 



4) . Publications: 

1. Specter, A. A., Fletcher, J. E., and Ashbrook, J. D.: Analysis 
of long-chain free fatty acid binding to bovine serum albumin by 
determination of stepwise equilibrium constants. Biochemistry. 
10: 3229-3232, 1971. 

2. Specter, A. A., Ashbrook, J. D., Santos, E. C., and Fletcher, J. E. 
Quantitative analysis of free fatty acid uptake by mammalian cells : 
Laurie acid and human erythrocytes. J. of Lipid Research , 1972. 
(in press) . 

3. Ashbrook, J. D.: A user's guide to MODELAIDE on the IBM system 
360/370. Bethesda, Maryland, Division of Computer Research and 
Technology, 1972, 16 pp. 



85 



Serial No. DCRT 3.6 

1. Laboratory of Applied Studies 

2. Office of the Chief 

3 . Bathe sda 

PHS-NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Simulation and Modelling in Neurophysiological Research 

Previous Serial Number: Same 

Principal Investigators: James A. Mortimer CPi"oject A) 

Erik Pottala (Projects B and C) 

Co -Investigators: E. V. Evarts, LNP, mm (Project A) 
T. R. Colbum, IR, NIMi (Project B) 
R. E. Burke, li^LC, NINDS (Project B) 
R. G. Justiss, CSL, DCRT (Project C) 
M. Bruce, CSL, DCRT (Project C) 

Man Years: Professional - 2.2 

Other - 0.1 ' . 

Project Description: 

A. A Functional Model and Experimental Study of Mammalian Cerebellar 
Cortex 

Objectives: 

1) To evaluate alternative hypotheses of cerebellar function through 
study of the electrophysiological activity of the monkey cere- 
bellum in relation to motor activity; 

2) To relate the physiological findings of this study to known cere- 
bellar circuitry by means of a computer simulation model ; 

3) To develop a new functional model for the mammalian cerebellum 
suitable for computer simulation. 

Progress during FY 72: 

Major attention during this year has been directed toward collecting 
physiological data that can serve as a basis for developing more 
realistic models of the cerebellum. In the initial set of experiments, 
the functional roles of the climbing fiber input to the cerebellum 
and of the cerebellar nuclei have been examined. 



86 



Past efforts at identifying the temporal sequence of activity taking 
place in the cerebellum in association with movements have been hampered 
by the difficulty of finding those neurons whose activity was related to 
a particular movement. This sampling problem has been avoided in the 
present investigation by studying the spike activity of single Purkinje 
and cerebellar nuclear neurons in relation to the sterotyped startle 
response evoked by a sudden auditory stimulus. Because of the synchroni- 
zation of activity in entire populations of cerebellar neurons brought 
about by this stimulus, the complexity of the cerebellar response is 
markedly reduced, and it is possible to examine the temporal relationship 
of each component of the cerebellar response to the time course of the 
evoked motor activity. Using this technique several existing hypotheses 
of cerebellar operation have been examined. In this series of experiments, 
the following major findings were made: 

1) Qianges in the rate of firing of cerebellar nuclear cells take place 
earlier than those in Purkinje cells. This finding suggests that a 
major amount of information processing in the cerebellum must occur 
in the nuclei. 

2) Climbing fiber responses occur in a time-locked relationship to the 
startle response, as expected from the hypothesis that their occur- 
rence is related to phasic motor activity. 

3) In conjunction with recent ablation studies, the finding that nuclear 
cells are activated prior to the earliest changes in electromyographic 
activity suggests that these neurons play a role in the initiation as 
well as the control of movement. 

These findings have been reported in a preliminary paper. 

Proposed Course: 

The work, to date, has been primarily concerned with the first project 
objective. During the next year, the computer model of the cerebellar 
cortex will be expanded to include the cerebellar nuclei and reprogrammed 
for more efficient operation. Using this model, the data collected in 
the series of experiments described above will be examined to determine 
the role of the various neuronal elements in generating the observed 
cerebellar responses. Collaborative use of the model with neurophysio- 
logists outside of NIH is also planned. In addition, new physiological 
experiments directed toward understanding the role of cerebellum in the 
initiation and control of movement will be carried out. 

Publications: (see end) 



87 



Hardware Neural Modelling 

Objectives: 

The objective of the hardware neural modelling study is to investigate 
small neural nets through the use of physiologically realistic hard- 
ware neural models which incorporate a distributed input system 
(analogous to a dendritic net) and are able to simulate action potentials 

Progress during FY 72: 



( 



The model has given insight into the manner in which an action potential 
can modify the shape and duration of post-synaptic potentials and the 
spatial -temporal interactions among such potentials. A preliminary 
study of the steady-state input -output firing characteristics of the 
model were presented in part at the 24th Annual Conference on Engineering 
in Medicine and Biology. The full study has been submitted to the 
IEEE Transactions on Biomedical Engineering. 

The neural model has been successfully interfaced to the LAS Computer 
CMAC-16) . 

Proposed Course: 

With the data acquisition and display capabilities of the model under 
computer control, it will be possible to study the model's behavior ^ 
to increasingly complex stimulus patterns. These studies will include I 
both complex sub-threshold neural responses and small neural net 
behavior . 

C. MAC-16 Research System Development (the LAS Computer) 

Background : 

A mini -computer has a place as a research tool if it has a variety of 
interfaces to control different types of peripheral devices and if it 
:/ has an easy-to-use Research Programming System which eliminates the 

need of the researcher to know the fine points of these various peri- 
pherals. To obtain these desired hardware and software systems is 
the objective of this project. 

Progress during FY 72: 

1 . Hardware : 

The following devices have been successfully interfaced to the MAC-16 
Computer: 

a. The Marquette Cart (to collect analog ECG tapes) 

b. The Honeywell 7600 Analog Tape Transport 

c. The neural control panel 

d. General purpose switch- filter network 

88 



2 . Software : 

In order to control the above peripheral and have a "usable" system the 
following software systems have been implemented. 

a. An analog to digital conversion system to service the Marquette tapes 
which provide data for the ECG analysis programs on the 360. 

b. An interactive analog to digital conversion system to perform general 
A/D conversion and provide data that can be further analyzed on either 
the 360 or PDP-10. 

c. A neural control system to provide for multiple input and output 
control of the previously mentioned neural model. 

d. A general research programming system consisting of a monitor program, 
a digital tape transport control program, and a graphics control pro- 
gram. The system is resident on the LING Tape System and eliminates 
the need for paper tape programs to control the peripherals. 

Proposed Course: 

Software: Continued development of programs to simplify the use of the MAC-16 
system for the general user. 

Hardware: To interface a real time spectrum analyzer to the MAC-16 and to 
investigate the feasibility of interfacing the MAC-16 to the 360 
or PDP-10. 



Publications: 

1/ 

J. A. Mortimer and E. V. Evarts: Latency differences m cerebellar Purkmje 
and nuclear cell activity in association with startle responses. Federation 
Proceedings. 31: 378, April, 1972. 



89 



c 



i 



July 1, 1971 through June 30, 1972 



PUBLIC HEALTH SERVICE - NATIONAL INSTITUTES OF HEALTH 
DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 

Summary of Branch Activities 1. DCRT - 4 



Serial Number 

2. COMPUTER SYSTEMS LABORATORY 3. Daniel Syed 

Acting Chief 

The Computer Systems Laboratory (CSL) concentrates on data system problems 
that do not lend themselves to solution through use of DCRT's central computer 
facility and require instead, the development of peripheral computer systems 
or special purpose hardware. CSL designs and implements computer based data 
systems for any promising application at NIH, even though during our first 
few years we have concentrated on the areas of laboratory automation and the 
delivery of health care. 

A specific area which promises potential benefit by the application of the 
computer technology is identified; the way in which the computer is to be used 
is determined, the hardware and software aspects of the system are designed 
and implemented, and the users instructed in the use of the system. Such work 
requires a mul tidiscipl inary group effort and requires two to four years from 
inception to completion. 

Because of the large number and diversity of experiments and instruments and 
the unique combinations of requirements, NIH represents a unique environment 
for the application of computers. There has been no experience in the appli- 
cation of computers in such an environment, so at the outset, one can only 
speculate as to the potential benefits of "computerization." We have there- 
fore engaged in what amounts to experiments to determine the benefits of com- 
puters in several different kinds of research laboratory and medical care 
environments. These trials, to determine whether or not the computer will 
really facilitate better research, involve a sizeable investment in capital 
and development manpower. 

Within the next year or two some of the systems which we have developed will 
be operating sufficiently smoothly to assess their value. Only then can the 
true cost-benefit relationship for these systems be determined. A lengthy 
period is necessary, not only because of the size of the development effort 
but also because it takes time for the scientists to get accustomed to using 
the system. In some cases they may need to adjust their experimental pro- 
cedures and the way they collect and use data to get the maximum benefit from 
their new tool, the computer. NIH is providing leadership and innovation in 
this type of computer application; in evidence of which, the U.S. Geologic 
Survey, Merck, Sharpe & Dohme and the FDA, among others, have been here to 
look over some of these computer systems and are considering copying them. 



91 



Computers in the Laboratory 

The laboratory automation work, to date, falls in two categories: the 
"computerization" of biochemistry laboratories, and the "computerization" of 
psychological studies. In the first category there is the computer system for 
NIDR, one for NIAMD, one for the NHLI's Mass Spectrometers, one for the NICHD's 
Gerontology Research Center in Baltimore, and one for use with the PSL-DCRT 
NMR Spectrometer. 

Although the systems are in many ways different, they are all used to collect 
data directly from instruments, to calibrate, to format, to perform calculation 
on the data, and to produce analyzed experimental results for the bench scien- 
tist, during, or immediately after, his experiment. The processed data is 
then available in a form that should be easy to use by the experimenter, for 
example, as plotted results or graphical results displayed on a CRT. Gener- 
ally, the system collects data from several instruments simultaneously so the 
marginal cost per user, or per experiment, is minimized. The systems provide 
the scientist with analyzed data at the time he is conducting his experiment, 
thus allowing him to modify his experiment while it is still set up and his 
experimental samples are still available. It allows him to do some experi- 
ments that he could not do before, for example, stimulus-response experiments 
using implanted micro electrodes can be performed, the stimulus changed and 
more data collected and analyzed before the cell in which the electrode is 
implanted dies. These computers also allow him to analyze more data in more 
complex ways than was ever possible before. The state of progress of these 
systems is reported in the individual progress reports 4-1, 4-2, 4-3, 4-4, 4-5. 

In the second category is the NIMH Building 10 system to be used to facilitate 
psychological studies. In these studies the patients interact with the com- 
puter. Here also, the computer is shared among a number of concurrent studies 
--each scientist can use the computer simultaneously, yet independently, of 
the others. The system serves to collect data, control and modify game situ- 
ations and control and format displays of various kinds. 

Computers in the Delivery of Health Care 

There are three projects in this Laboratory relating to the delivery of health 
care. The first one (4-7) is an effort to bring easy to use, economical and 
reliable computer services to the practicing physician. He will be able to 
use a ubiquitous terminal such as a standard telephone, to access commercial 
computer systems for assistance in such activities as diagnosis and therapy 
planning. This system is designed with the community physician in mind, as 
well as NIH clinical requirements. The second and third projects relate to 
the care of critically ill patients. In these two projects (4-8 & 4-11) we 
have concentrated on using the computer in cardiovascular studies with NHLI, 
and in working toward a computer system to assist in an intensive care envi- 
ronment. The details of progress in these projects can be found in the 
Individual Project reports. 

Most of these projects began prior to this reporting year. Most represent 
collaborative work with other Institutes. Attached is a table which shows, 
graphically, the CSL personnel effort put into these projects over the last 



92 



3 years, the capital expenditure required for these projects and the source of 
these funds. The capital expenditures portion of the table has one entry for 
equipment which includes both commercially purchased equipment and the cost of 
components for the equipment developed in this Laboratory, and one entry for 
central facility computer charges expended on the project for computer program 
development. The personnel effort includes that required for problem analysis, 
system engineering, equipment design engineering and programming, all of which 
is done by professionals in the field of engineering, mathematics, physics, 
chemistry, medicine, and computer science. 



93 



CSL MANPOWER CAPITAL 

(Man-Years) Equipment Computer 

Project # Name FY 69-70 71 72 Source Amount FY Charges 

DCRT 4-1 NIDR 10 2 3 NIDR 100(K$) 68 



4-2 NIAMD 10 

4-3 Mass Spec. ■ 4 

4-4 Gerontology 4 

4-5 NMR 1/2 

4-6 NIMH 1 1/2 

4-7 Med. Telecomm. 12 

4-8 Cardio-Vascular 8 

4-10 Consultation 3 1/2 

4-11 Intensive 

Care Unit 







DCRT 


150 


69 








DCRT 





70 


12(K$) 






DCRT 





71 


1 






NIDR 


115 


71 









DCRT 





72 









NIDR 





72 


1 


4 


3 1/2 


NIAMD 


180 


68 








DCRT 


25 


69 








DCRT 


20 


70 


3 






DCRT 


2 


71 


6 






DCRT 





72 


3 


2 


2 1/2 


DCRT 


76 


69 








DCRT 


10 


70 


1/2 






DCRT 


2 


71 


1 






DCRT 


14 


72 





2 


2 1/2 


NICHD 


111 


69 








NICHD 


33 


70 









NICHD 





71 









NICHD 





72 





1 1/2 2 


DCRT 


68 


70 









DCRT 





71 









DCRT 


5 


72 


2 


4 


2 


NIMH 


243 


70 









DCRT 





71 


5 






DCRT 





72 


6 






NIMH 





72 


5 


6 


2 1/2 


DCRT 


125 


69 








DCRT 


112 


70 


12 






DCRT 


16 


71 


29 






DCRT 





72 


6 


4 1/2 


1 1/2 


DCRT 


20 


69 








DCRT 


21 


70 


1 






DCRT 


11 


71 


2 






DCRT 





72 





1 


1 


DCRT 





71 









DCRT 





72 





1 


6 


DCRT 


370 


71 









DCRT 


20 


72 


10 






NHLI 


15 


72 






94 



Serial No. DCRT 4-1 
Computer Systems Laboratory 
Systems Design Section 
PHS-NIH Bethesda 



Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Computer System for NIDR 

Project Leader: Daniel Syed 

Objectives : 

The use of the computer as an adjunct to laboratory procedures in the 
National Institute of Dental Research is intended to both augment and 
expedite the biochemistry and neurophysiology research programs of that 
Institute. Specifically, the computer system will modernize techniques for 
acquiring data from laboratory instruments and will provide for the real- 
time analysis of data from selected experiments. Ultimately, it is planned 
to use the computer to control experiments in real-time. 

Background: 

Subsequent to a detailed requirement study started in February 1967, a pro- 
cess control class computer system was designed and specifications generated. 
Open competition procurement resulted in the selection of a Honeywell DDP-516 
computer which was purchased in June 1968 and installed in the Dental Insti- 
tute, Building 30, in July 1969. Honeywell supplied system software has been 
modified to permit operation in a mul ti programmed mode and the system is now 
capable of acquiring and analyzing data acquired simultaneously from amino 
acid analyzers, scintillation counters and neurophysiological experiments. 
In addition to the generation of applications programs, user oriented execu- 
tives have been implemented for use with neurophysiological projects and 
experiments utilizing scintillation counters. An experimental link to the 
central PDP-10 computer has been completed. 

FY-72 Activities: 

During the past year four additional instruments were interfaced to the 
Honeywell 516 computer. These included a spectrophotometer, a gas chromato- 
graph, an X-ray diffractometer and a ph meter. The system is now supporting 
around the clock data acquisition and analysis from these instruments. The 
continuous mode of operation for many of the instruments connected to the 
computer presented serious problems for program development and checkout. 
To circumvent these problems the system was expanded to include a Honeywell 
316 processor. A communication link to the central facility IBM 370/360 
computers has been implemented. Engineering redesign and modification of 
several instrument interfaces has resulted in a substantial improvement in 
the overall performance of the system. Various utility programs and refine- 
ments to the operating system were made to facilitate use of the system by 
the research investigators. 



95 



Future Efforts: 

Efforts during the coming year will be concentrated on developing user 
oriented software to allow use of the 516/Central Facility communication 
link. Implementation of software to permit the Honeywell 516 and 316 com- 
puters to share peripheral equipment is also planned. An NMR spectrometer 
and spectropolarimeter will be interfaced to the Honeywell 516 computer. 
Continuing efforts will be required to simplify use of the system by the 
research investigators. 



96 



Serial No. DCRT 4-2 
Computer Systems Laboratory 
Laboratory Applications 
PHS-NIH Bethesda 

Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Computer System for NIAMD 

Project Leader: Marvin Shapiro 

Objectives: 

The NIAMD computer located in Room 110, Building 2, is a real-time data 
acquisition system which has been in operation since March 1970. It is 
being used as a research tool by scientists in a number of ways. Data 
collection is done quickly and accurately with the computer and is then 
available in the computer for further processing. This represents a signifi- 
cant reduction in the time required to complete calculations and to plot 
results over previous methods. These improvements, plus the ability to use 
the computer for feedback control of some experiments, should result in a 
significant change and improvement in the experimental procedures being used 
by many scientists in NIAMD. 

Background: 

A computer system for real-time data acquisition was designed in early 1968 
and a system meeting the requirements (a Honeywell DDP-516 computer) was 
ordered in June of 1968. The system was delivered in July 1969. NIH work 
on interfacing instruments to the computer and on software modifications to 
the monitor supplied by Honeywell began in July 1968. The system became 
operational for routine data collection in March 1970, at which time four 
instruments were interfaced to the computer. 

Fiscal Year 1972 Activities: 

Major improvements were made in the data acquisition operating system. Capa- 
bilities for displaying selected portions of data files were added. The 
library of systems and data analysis programs continued to be expanded. 

Two small computers, a PDP-8 controlling a diffractometer, and a Raytheon 704 
used in retinal research, were interfaced to the 516 and software was devel- 
oped for transferring data and programs between computers. Hardware for 
interfacing the light stimulus experiments of Dr. Hagins to his 704 computer, 
both for collecting data and controlling the experiments, was completed and 
checked out. 

Initial work connecting the 516 to the central 360 computer system via phone 
1 ines was completed. 

In fiscal 1971 two more instruments, a Computer of Average Transients (CAT) 
and an analytical ul tracentrifuge, were interfaced to the computer and 
incorporated into the software operating system. 

97 



Future Efforts: 

During the coming year work to be done will include the following: 

1. Interfacing of more instruments, including a CARY118 spectrophotometer. 
A considerable amount of software will have to be written to support 
the 118, which will require feedback control of temperature over 
relatively long periods of time. 

2. A package of programs for support of the 704 computer will be developed, 
so that assembling, program editing, and data analysis can be done thru 
the 516 and the results transmitted over the direct line to the 704. 

3. The initial connection to the central NIH computer system via telephone 
lines will be completed. This will enable data files to be transferred 
between the computers. Eventually the capabilities will be increased 
to allow initiation of 360 programs from the 516. 

4. The development of more application software for individual instruments, 
notably the ul tracentrifuge. 



98 



Serial No. DCRT 4-3 
Computer Systems Laboratory 
Laboratory Applications 
PHS-NIH Bethesda 
Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Mass Spectrometer Related Work 

Project Leader: Marvin Shapiro 

Objectives: 

The broad objective is the continued support of research involving the col- 
lection and use of mass spectrometer data. The work has thus far been done 
in collaboration with scientists in the Laboratory of Chemistry, NHLI. 

Background: 

In 1968 work began on interfacing a PDP-8I computer to an MS-9 high resolution 
mass spectrometer in the NHLI. In 1970 a low resolution, LKB mass spectrom- 
eter was connected to the computer. Since early FY-71 the system has been 
used on a routine basis to obtain medium resolution mass spectra of compounds. 

FY-72 Activities: 

Chiefly, four projects were worked on: (1) The development of the computer 
system described has continued with the software addition of a plotter pro- 
gram and the completion of a system for collecting low resolution data. 
(The latter work was done by Dr. Redder of NHLI.) Additional hardware has 
included a connection to the PDP-10 via a modem and a mass marker identi- 
fication signal (under development) for masses above ten. (2) A small mass 
spectrometer-computer system was purchased from Finnigan Corporation for the 
collection and analysis of protein sequencing data. The system was installed 
in December 1971, and in April 1972 began being used on a routine basis. CSL 
work on this system has thus far consisted of help in writing specifications 
for procurement. (3) Work has begun on using low resolution mass spectrom- 
eter data for the classification of drug activity. Data has been obtained 
from standard libraries of mass spectrometer data and is being collected by 
Dr. Milne of NHLI. Classification of drug activity has been done by 
Dr. Guarino of NCI and Dr. Ting of CSL. Programs have been written using 
cluster analysis techniques and learning machine approaches. (4) A version 
of the poison identification program was written for the voice-answerback 
computer SEL and frequent demonstrations have been given. The user dials 
the computer, then, on request, inputs (via the touch-tone telephone buttons) 
the 5 largest peaks shown by the mass spectrometer run of the poison. Then 
the computer searches its poison file and speaks back the name of the poison 
identified. 



99 



Future Efforts: 

Future work lies mainly in areas (2) and (3) above and in the development 
of programs to file and retrieve the low resolution data collected under 
(1). Some programming will be required to tailor the Finnigan system more 
specifically to the needs of the problem it is being used for and to develop 
display software. 

The drug classification work is "jery long range in nature and will require 
the continued collection of drug data, the development of good activity 
classification for the drugs, and the continued development of classification 

algorithms. 



100 



Serial No. DCRT 4-4 
Computer Systems Laboratory 
Project Development Section 
PHS-NIH Bethesda 

Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Computer System for Gerontology Research Center 

Project Leader: Perry S. Plexico 

Objectives: 

The computer system at the Gerontology Research Center (GRC), NIHCD, in 
Baltimore, is used to accommodate various on-line control and data acqui- 
sition experiments and off-line data processing tasks in support of research 
by the Center into the nature of the ageing process. The research efforts 
are multidisciplinary and encompass the techniques of biochemistry, molec- 
ular biology, physiology, and psychology; hence, the applications for which 
the computer is used are widely varied. 

Background: 

A computer system meeting the research needs of GRC and based upon system 
design specifications developed by the Computer Systems Laboratory, was 
procured from the Raytheon Company, and accepted by the Government in May 
1970. Since that time, the system has been used by GRC research investi- 
gators for off-line processing and for on-line psychological testing and 
problem solving experiments. During the same period, members of the CSL 
staff have been engaged in the development of software to automate a 
glucose/insulin infusion control experiment. This experiment is a feedback 
control process in which the computer controls the infusion of glucose into 
a subject's bloodstream based upon the results of monitoring blood glucose 
level, the objective being to achieve a better understanding of the under- 
lying causes of diabetes and the effect of ageing on glucose tolerance. 

FY-72 Activities: 

Software to support the infusion experiment was completed during the fall 
of 1971; however, testing was delayed due to malfunction of the contractor 
supplied operating system which have since been corrected. All controlled 
testing has now been completed and the software released to GRC research 
investigators for partially controlled testing and evaluation during actual 
experiments. 

Future Efforts: 

Upon completion of the final in-use testing of the glucose/insulin infusion 
software, any problems discovered during this phase will be corrected. The 
software will be released for routine use by GRC personnel and CSL involve- 
ment in this project terminated. 



101 



Serial No. DCRT 4-5 



Computer Systems Laboratory 
Processor Design Section 
PHS-NIH Bethesda 

Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Pulse Nuclear Magnetic Resonance (NMR) Spectroscopy for 
Carbon-13 and Proton Studies 

Project Leader: Victor Col burn 

Objectives : 

The Physical Sciences Laboratory, DCRT, in collaboration with the NIAMD, 
plans to use a pulsed NMR spectrometer in conjunction with a small scale 
on-line computer system for the performance of high resolution Fourier 
transform spectroscopy. The advantage of Fourier transform spectroscopy, 
as compared to conventional frequency sweep techniques, is that signifi- 
cant improvement in signal-to-noise and/or in shortening of experiments 
run times can be realized, thus making feasible a whole new range of 
research experiments. 

Background: 

As a result of the Computer Systems Laboratory system study on NMR Fourier 
transform spectroscopy, a computer system was configured, proposed, and 
procured in June 1970 for carbon-13 resonance work. At the same time, BEIB, 
DRS, undertook design and implementation of a radio frequency pulse system 
to operate in conjunction with an existing NMR sweep spectrometer in NIAMD. 
The overall system is being tailored for carbon-13 resonance work at the 
pulse frequency of 55 megaHertz. 

FY-72 Activities: 

CSL completed performance tests on the R-704 computer system and also 
designed, fabricated and installed additional new interface hardware for 
filtering the raw analog data signals and for controlling the pulsed-NMR 
data handling procedures within the computer system. CSL completed and 
installed a system of eight computer programs for data acquisition and 
averaging and for recording of resulting time domain data tables on magnetic 
tape. BEIB installed their r-f pulse system on the NMR spectrometer. All 
necessary interconnections were completed and checked out. The overall 
NMR spectrometer system is currently operational for the acquisition, aver- 
aging and recording of pulse-NMR time domain data. These data are currently 
being converted to plotted frequency spectrum by Fast Fourier Transformation 
on the DCRT PDP-10 computer and Calcomp plotter combination. 

Additionally, it was decided in early FY-72 to further modify the same basic 
NMR sweep spectrometer to also accommodate pulse proton resonance work at 
the corresponding r-f pulse frequency of 220 megaHertz. Such a modification 
is now commercially available which includes an r-f pulse system for 220 



102 



megaHertz and a small computer for data accumulation and Fast Fourier Trans- 
formation. A study by CSL indicated that it would not be ecnomically fea- 
sible to attempt any other style of modification, such as one involving a 
single computer to serve both systems; therefore, the commercial modification 
was ordered on 3/13/72 and is now due for delivery by 9/17/72. 

Future Efforts: 

With regard to the computer aspects of the Carbon-13 system, work will con- 
tinue on the development of the software system to provide the following 
features on the R-704 computer at the pulse-NMR spectrometer site: (1) Fast 
Fourier Transformation for various large size data tables, (2) CRT display 
for both data acquisition and frequency spectrum data, (3) Hard copy output 
for frequency spectrum plots, (4) Various improvements for more efficient 
overall system performance. 

Further attention will go to integrating and coordinating the pulse proton 
modification within the overall NMR Spectrometer system. 



103 



Serial No. DCRT 4-6 
Computer Systems Laboratory 
Systems Design Section 
PHS-NIH Bethesda 



Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Computer System for NIMH 

Project Leader: Daniel Syed 

Objectives : 

This project involves implementation of a general purpose computer system 
for support of research studies which are being conducted in the National 
Institute of Mental Health at NIH. The system provides a facility for on- 
line data acquisition, real-time control and post-experiment data analysis 
and storage for researchers in the fields of psychology and neurophysiology. 
Currently, investigators are using the system to study Concept Formation, 
Evoked Response, and Perception. The system will be used to support investi- 
gations which relate electroencephalographs and other physiological variables 
to clinical diagnosis and classification of NIMH patients. 

Background: 

CSL conducted studies of NIMH requirements during calendar 1967. Funding 
problems limited initial procurement to 70% of the intended total system. 
The computer system was installed on June 30, 1970 and officially accepted i 
by NIMH in October. By July 1971 several experiments had been implemented 
and a mode of operation established. Engineering support and some pro- 
gramming was provided by NIMH personnel. CSL was responsible for system 
software design and implementation. 

FY-72 Activities: 

During the first six months of fiscal year 1972 CSL effort was mainly concen- 
trated on removing system errors and providing improvements to the executive 
software. Limited application software support was provided to real-time 
users. Typical system enhancements included the provision of software linkage 
paths between real-time and batch programs and the modification of several 
input/output handlers so that fault conditions no longer destroyed overall 
system effectiveness. By mid January 1973 the system had achieved a satis- 
factory degree of reliability. From mid- January until mid-May CSL was able 
to support this system on an on-call basis only. 

Future Effort: 

The NIMH system now supports five real-time users and a number of off-line 
analysis programs. During the coming year it is planned that CSL will assist 
in the implementation of several additional real-time users. No modification 
to the system should be necessary to accommodate these users. It is also 
planned to introduce a number of additional system modifications aimed at 
advancing the effectiveness and reliability of the system. 



104 



Serial No. DCRT 4-7 
Computer Systems Laboratory 
Project Development Section 
PHS-NIH Bethesda 

Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Medical Telecommunications 

Project Leader: Perry Plexico 

Objectives : 

The objectives of this project are: (1) to facilitate use of the computer 
for the practicing clinician, medical research physician and supporting 
paramedical personnel by developing convenient economical and reliable 
methods for remote access to computer-stored information and computations, 
and (2) to develop useful computer programs to assist the health profes- 
sional in patient-care decision-making. 

Background: 

While the widespread use of time-sharing computer systems has produced the 
potential for applying computer technology to some of the problems of 
health care delivery, the high cost, large size, and lack of wide avail- 
ability of standard terminal devices render them unsuitable for clinical 
applications. These factors led to the establishment of a project to 
investigate the use of telephones as computer terminals. As a result, a 
computer system has been developed to permit inquiries and data to be 
entered into the computer from the keyboard of any standard pushbutton tele- 
phone and to provide voice response to the user over the same telephone via 
computer driven audio response units. 

FY-72 Activities 

The design goals originally established by this project were that of (1) a 
network of computers consisting of a number of Government owned and com- 
mercial time-sharing services on which medical applications programs would 
be maintained, and (2) a single communications processor for accepting 
touchtone inputs, producing voice outputs, and selecting and dialing time- 
sharing services. The concept was to allow automatic switching from one 
time-sharing service to another so as to provide adequate reliability, even 
when one or more time sharing systems were inoperative. Development of a 
system meeting these requirements was completed late in 1971. Unfortunately, 
however, the long response time delays resulting from dialing time sharing 
systems and low speed data transmission between computers, as well as the 
problems of the communications processor adapting to the various input/ 
output formats of multiple time-sharing computers far outweighed the positive 
factors mentioned above. On this basis, work has begun on what is expected 
to be a final development phase of this project. During this phase a multi- 
programming operating system will be implemented on the computer currently 
used as a communications processor. This will allow both applications pro- 
grams and telephone terminal support to reside on the same computer. This 

105 



effort will correct the operational problems associated with the current . 
system and will have the added advantage of making the resources of the I 
voice response computer available for other activities. 

Future Efforts: 

The current developmental effort is expected to be completed during the 
summer of 1972 at which time the system will be made available to interested 
members of the NIH staff and to the local medical community. The future ^ 
direction of this project will depend upon the comments, criticism, and ^ 
suggestions resulting from this trial use. 



106 



Serial No. DCRT 4-8 
Computer Systems Laboratory 
Systems Design Section 
PHS-NIH Bethesda 



Project Report 
July 1, 1971 through June 30, 1972 



Project Title: Cardiovascular Studies 
Project Leader: Daniel Syed 

Objectives: 

The beat-by-beat analysis of electrocardiographic, blood pressure and other 
biomedical waveforms can provide great insight into the pathophysiology of 
the cardiovascular system. In collaboration with the Cardiology and Sur- 
gical Branches of NHLI, this project attempts to monitor and analyze these 
physiological variables through the use of recent developments in digital 
computer technology. 

Background: 

Software that allows the evaluation of electrocardiographic arrhythmias and 
S-T segment elevation/depression measurements on a beat-to-beat basis has 
been developed. Software systems have been implemented for the quantifica- 
tion of myocardial function by analysis of the motion of the heart border 
and the Mason Index analysis of the left ventricular pressure waveform. 

Special purpose hardware preprocessors have been developed to allow the 
analysis of cardiac arrhythmias and the detection of upstroke and dicrotic 
notch in the aortic pressure waveform. An optical isolation technique 
incorporated into the preprocessors to provide for patient safety was also 
implemented as a versatile in-line coupler for general purpose signal iso- 
lation applications. A finger-tip ECG acoustic coupler was developed to 
allow an ECG to be transmitted to a physician from any remote location via 
a standard telephone. 

An experimental study was undertaken in the animal laboratory which will 
allow the analysis of several arterial pulse contour methods for the 
estimation of stroke volume. An ultrasonic doppler flowmeter was utilized 
during this study in an attempt to obtain peripheral flow data which will 
allow the correction of errors introduced by changes in peripheral vascular 
resistance. 

FY-72 Activities: 

During FY-72 software has been developed in support of a treadmill exercise 
protocol contained within NHLI's Type II Hyperlipoproteinemia Study. S-T 
segment elevation/depression measurements and parameters derived from an 
arterial pressure waveform are computed periodically from tape recordings 
made during each treadmill session. Additional ECG software allows the 
statistical analysis of consecutive R-R intervals in an effort to develop 



107 



techniques for the differential diagnosis of atrial fibrillation and 
digitalis toxicity. 

A left ventricular diameter study was conducted in which ultrasonic diameter 
measurements were obtained along with LV pressure and aortic flow signals. 
Computer analysis of LV geometry was performed along with a simultaneous 
computation of Mason Index parameters and traditional quantifiers of LV 
function. 

A program was developed to allow a specialized analysis of the phonocardio- 
gram. The ball-in-cage heart valve prosthesis has been found to fail as a 
result of a crack or distortion occurring in its ball. A technique to detect 
such ball variance, which is being evaluated, consists of computing the stan- 
dard correlation coefficient between consecutive first and second heart 
sounds in a phonocardiogram. An interactive graphics terminal with a 
joystick-controlled cursor allows an investigator to select the portions of 
the heart sounds which are to be correlated. 

A series of programs has been written to allow the detailed analysis of 
data collected in the stroke volume estimation study. The result of the 
computer analysis will be a comparison of several proposed stroke volume 
estimating functions with each other and the electromagnetic aortic flow- 
meter. 

Hardware preprocessor development proceeded with the completion of a digital 
algorithm for the detection of ventricular premature contractions. Prelimi- 
nary evaluation of this development is encouraging. Other improvements have 
increased the accuracy and reliability of the family of ECG preprocessors. 

The finger-tip ECG acoustic coupler has been redesigned to include a 12 
button touch-tone pad. This modification allows a patient to communicate 
via touch-tone codes with a computer for unattended acquisition and analysis 
of an ECG. Such a prototype ECG analysis system has been developed utilizing 
the SEL-810B voice answerback computer system. 

Future Efforts: 

Further development of software algorithms will be carried out as research 
goals dictate. Software and hardware algorithms for biomedical signal 
analysis will be adapted for use in the intensive care unit computer system 
(Project DCRT 4-11). Final phase of stroke volume study will begin with 
the analysis of experimental data. 



108 



Serial No. DCRT 4-10 

Computer Systems Laboratory 

NA 

PHS-NIH Bethesda 

Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Computer Systems Laboratory Consultation 

Project Leader: Alan Demmerle 

Objectives: 

This project incorporates a number of consultation activities directed 
toward providing requesting Institutes with guidance in any aspect of 
computer systems design and use. 

Background: 

The project seeks to afford assistance to the intramural and extramural 
programs of all Institutes. However, to date, consultative activities have 
been directed primarily toward the support of the data management portions 
of large dollar value, long term contracts and grants sponsored by the 
NHLI. During collaboration with NHLI, efforts have included providing 
contractor guidance, designing a complete system or simply offering advice 
to NHLI contract office. 

FY-72 Activities: 

During the past year effort was concentrated on the support of two major 
NHLI projects: the Myocardial Infarction Program and the Blood Resources 
Program. 

Future Activities : 

Future efforts in this project are determined by ad hoc requests from the 
Institutes. 



109 



Serial No. DCRT 4-11 
Computer Systems Laboratory 
Systems Design Section 
PHS-NIH Bethesda 

Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Intensive Care Unit Computer System 

Project Leader: Daniel Syed 

Objectives : 

This project involves development of a computer system for intensive care 
monitoring that performs the following functions: (1) vital sign monitoring 
with feedback to clinicians and nursing staff, (2) statistical analyses, 
trend identification and other analyses performed on historical data, and 
(3) eventual computer control of selected patient variables. Performed in 
collaboration with the Surgical Branch of NHLI, on-line data will be obtained 
from the post-operative recovery beds of NHLI. 

Background: 

In 1969 members of the CSL staff designed a computer system to be used for 
intensive care patient monitoring at five universities funded for Myocardial 
Infarction Research by NHLI. This system is being modified, and installed 
in the NIH Clinical Center environment. Procurement of the computer and 
preparation of the site was initiated during 1971, and some first phase goals 
were developed. 

FY-72 Activities: 

CSL began development of systems software modules during the Summer 1971 in 
anticipation of computer delivery in November 1971. Extensions to the oper- 
ating system for handling the color display terminal and the analog to digital 
converter were completed prior to delivery. A simple system capable of per- 
forming arrhythmia analysis on one ECG lead was written to provide the basis 
for acceptance testing of the computer hardware. The hardware was accepted 
in February 1972. Development of further system software to handle the multi- 
terminal graphics system and hardware signal preprocessors, as well as soft- 
ware to do analysis of arterial pressure, urine output, blood drainage and 
core temperature from one bed has been written. The physiological signals are 
transmitted from Building 10 to Building 12-A by Dataphone, and monitoring 
results are transmitted back to the display terminal in the recovery room. 

Future Efforts: 

The computer system will be moved to Building 10 when the computer room con- 
struction is complete, during the Summer, 1972. Effort will then be directed 
toward implementation of direct data acquisition from all beds and utilization 
of the more sophisticated permanent display system. Eventually, computer con- 
trolled infusion pumps will be implemented. 



110 



July 1, 1971 through June 30, 1972 

PUBLIC HEALTH SERVICE - NATIONAL INSTITUTES OF HEALTH 
DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 
Summary of Branch Activities 1. DCRT - 5 

2. PHYSICAL SCIENCES LABORATORY 3. Dr. G. H. Weiss 

I. OBJECTIVES 

The Physical Sciences Laboratory is devoted to the study o£ problems in 
physics and chemistry that relate to the biological sciences. Several 
disciplines are represented in the membership of the laboratory. These 
include applied mathematics, theoretical chemistry, and theoretical physics. 
Whenever possible the theoretical studies are performed in conjunction with 
experimental work, either in collaboration with workers in outside units, 
or by members of the Physical Sciences Laboratory working in other labora- 
tories at NIH. In addition to perfommig research of its own choosing, 
members of the Physical Sciences Laboratory provide consultation to other 
researchers at NIH on different topics in the disciplines represented in 
the Laboratory. These services are enumerated in the project reports. 

II. CURRENT LABORATORY PROGRAMS 
Summary 

1. We have made a systematic development of the singular perturbation 
method for linear parabolic partial differential eqviations to orders great- 
er than one. These equations describe the kinetics of a large class of 
biochemical separation procedures. Detailed applications have been made 

to effects of hydrostatic pressure in the ultracentrifuge, to diffusion 
broadening of peaks in pore gradient electrophoresis, and to the effects of 
nonuniform packing in analytical gel pore chromatography. 

2. An international synposium on numerical methods in chemistry was held 
at NIH in March 1972. Progress in methods of interpreting results of ultra- 
centrifugation and other biochemical separation methods was summarized and 
new problems and approaches were discussed by the participants. Other 
talks were given on the application of numerical methods to the determina- 
tion of chemical structure starting from either a quantum mechanical basis, 
or more enpirical laws for forces and torques . 

3. A new model of the helix-coil transitions in polypeptides has been 
developed, that is consistent both with the equilibrium Zimm-Bragg theory 
and with a variety of relaxation methods. The distinctive variation of 
dynamic properties with molecular weight is correctly accounted for by the 



111 



present theoiy, in contrast with other proposed theories. Progress has 
been made in developing accurate bounds on certain conformational properties 
of polymer molecules. These will allow more accurate specification of 
polymer structure from measured moments. 

4. A theory of the movement of bands of chemotactic bacteria has been 
developed and experiments were performed which verify major findings of 
the theory. A technique for the use of laser scattering in detecting 
spatially anisotropic motion and measuring electrophoretic mobilities has 
been developed and tested. 

5. The theory of van der Waals forces has been extended to take into 
account anisotropic and inhomogeneous dielectric properties. Approximate 
results have been obtained for the interaction of long thin molecules in 
solution. The theory has also been more intimately connected with available 
experimental data derived from reflection spectra. 

6. Analysis of data on head injured veterans of the Korean war has shown 
that the only reliable predictor of post -traumatic epilepsy is the number 
of fits within the first year post injury, i.e., that the degree of severity 
of injury, coma, and so forth are not good predictors of eventual epileptic 
fits or of the severity of such fits. Further work has shown that the 
only detectable effects of injury events on work status of these veterans 

15 years post injury is that a certain excess fraction (over controls) do 
not enter the work force. Those that do enter the work force are indisting- 
uishable from the control population. 

7. The NDDELAIDE program was used extensively in combination with NMR 
studies of protein interactions at the molecular level. It was found that 
the renaturation of staphylococcal nuclease is reversible, and that it 
occurs by multiple equilibria with different nucleation sites. Mathematical 
models have been developed that describe titration curves of proton resonances 
of imidazole in histidine, and estimation procedures were devised for the 
estimation of microscopic ionization constants. Carbon 13 Fourier trans- 
form NMR has been applied to study the formation of active enzyme conplexes. 



112 



Serial No. 5.1 

1 . Physical Sciences Laboratory 

2. Not Applicable 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Theory of Biochemical Separation Techniques 

Previous Serial Number: 5.1 

Principal LiArestigator: George H. Weiss, Ph.D. 

Other Investigators: None 

Cooperating Units: David Yphantis , Ph.D., University o£ Connecticut, 
Gary Ackers, Ph.D., University of Virginia, and 
David Rodbard, M.D. , Reproduction Research Branch, 
NICHD. 

Man Years 

Total: 0.9 

Professional: 0.8 
Other: 0.1 

Project Description: 

Objectives: 

To determine the physico-chemical effects influencing different 
biochemical separation systems such as ultracentrifugation, chromatography, 
and electrophoresis. To determine the quantitative significance of these 
effects . To devise numerical techniques for processing data from chemical 
separation procedures to determine properties such as molecular weight and 
diffusion coefficient. 

Methods : 

Numerical and analytical solutions to linear and nonlinear 
partial differential eqiaations. 

Major Findings: 

An investigation of hydrostatic pressure dependence of the 
diffusion constant in velocity sedimentation experiments showed that the 
only effect is a slight narrowing of the peak in concentration gradient 
without a shift in peak position. Hence the diffusion -free theory relating 
sedimentation coefficient to peak position can be used provided that 



113 



concentrations are sufficiently low that a linear theory can be used. 

A systematic development of singular perturbation theory for linear 
equations describing chemical separation systems has been conpleted, greatly 
inproving the accuracy near the peak position for small zone experiments. 

An exact solution and good approxijnations have been obtained for peak 
broadening in pore gradient electrophoresis. Close approximations have been 
obtained for the effects of nonuniform packing in analytical gel pore chroma- 
tography, showing that linear nonuniformity can be ignored at relative pack- 
ing differences ip to 20% across the column. 

The transient effects of finite equilibration times in gel pore chroma- 
tography have been studied and found to be negligible after approximately 
one minute, for typical system parameters in analytical work. 

Significance to Biomedical Research: 

A central problem in biochemistry is the separation of proteins. 
In the past chromatographic methods have been used for this purpose produc- 
ing crude and qualitative results. Presently electrophoresis and gel pore 
chromatography are being developed as analytical tools with some advantage 
over centrifugation. Analysis of these tools with a view towards the 
development of numerical methods for processing results of chemical 
separation experiments will lead to extremely accurate methods for character- 
izing inportant protein properties. 

Honors and Awards: None 

Publications : 

Dishon, M. , Weiss, G. H., and Yphantis , D. A. : A new approach to the 
effects of pressure on velocity sedimentation experiments. Journal of 
Polymer Science A2, 939-957 (1971). 

Dishon, M. , Weiss, G. H. , and Yphantis, D. A.: Kinetics of sedimenta- 
tion in a density gradient. Biopolymers 10, 2095-2111 (1971). 

Weiss, G. H., and Yphantis, D. A.: Pressure dependent diffusion in 
velocity sedimentation. Journal of Polymer Science A2 , 339-344 (1972). 

Weiss, G. H., Rodbard, D. : Diffusion dependent peak broadening in 
pore gradient electrophoresis. Separation Science 7, 217-232 (1972). 

Weiss, G. H. and Dishon, M. : Approximate solutions of chemical 
separation equations with diffusion. Advances in Chemistry (to appear). 

Weiss, G. H. and Dishon, M. : On the solution of linear chemical 
separation eqioations with weak diffusion effects. Separation Science 
(to appear) . 



114 



Serial No. 5.2 



1 . Phy s i cal Sciences Laboratory 

2. Not Applicable 

3. Bethesda 

PHS-NIH 
Individiial Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Theory of the Helix-Coil Transformation of Polypeptides in 
Solution 

Previous Serial Number: 5.2 

Principal Investigator: James A. Ferretti , Ph.D. 

Other Investigators: Robert Jemigan, Ph.D., George Weiss, Ph.D. and 
Julie Milstein, Ph.D. (NCI). 

Cooperating Units: National Institute of Arthritis and Metabolic Diseases 
Man Years: 



Total : 


1.1 


Professional: 


1.0 


Other 


0.1 



Project Description: 

Objectives: 

To elicit the molecular details of the kinetics of helix to 
random coil transitions in polypeptides. A further aim is the coup rehens ion 
of the action of proteolytic enzymes on polypeptides. 

Methods : 

A new model of the transition was developed. This model combines 
conventional polypeptide equilibrium statistics with random walk theory. 
Conparisons with less detailed theories were carried out. Also effects of 
molecular weight were investigated. 

Major Findings : 

Calculations with the new model yield average conformational 
lifetimes for each chain unit. This calculated distribution of relaxation 
times is consistent with a variety of measured relaxation properties. These 
include NMR, ultrasonic and electric measurements. The distinctive varia- 
tions of the NMR spectra with molecular size are explained only by the 
present model and a previous sinpler model proposed by several of the present 
workers. Sinpler models proposed by others are inconsistent with the 
e3<perimental results. 

115 



Significance to Biomedical Research: 

Most biological processes involving proteins are acconpanied by 
transitions between rigid helical regions and more flexible random coil 
conformations. Such transitions must occur in atLvo, whether during enzyme 
reactions or during protein synthesis. A detailed understanding of such 
simple analog models is required before the molecular details of protein 
conformational transitions can be conprehended . 

Honors and Awards: None 

P;±)lications: 

Ferretti, J. A., Jemigan, R. L. , and Weiss, G. H. : Nfodels of 
time -dependent behavior of polypeptides in the helix-random coil transition 
region. Macromolecular Preprint , International Congress of Pure and 
Applied Chemistry, Helsinki, 1972. 



116 



Serial No. ^-^ . , 

1. Physical Sciences Laboratory 

2. Not Applicable 

3. Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: I^lecular Mechanics 

Previous Serial Number : 5.4 

Principal Investigator: Robert L. Jemigan, Ph.D. 

Other Investigators: George H. Weiss, Ph.D. 

Cooperating Units: None 

Man Years: 

Total : 0.6 

Pro £es s ional : 0.6 
Other: 0.0 

Project Description: 

Objectives: 

The purpose is to predict dilute solution behavior of macro - 
molecules by means of detailed molecular models. 1) Equilibrium dimension- 
al properties are derivable from intra-chain distance distribution functions 
(d.f.). Even moments of distance are calculable for realistic models. These 
yield approximate d.f. 's which can in turn be applied to obtain bounds to 
other distance related functions. 2) Non-equilibrium properties can be 
treated in terms of time correlation functions. A general treatment of 
these with a realistic molecular model is desired. 

Methods : 

Experimental information on small molecules is combined with a 
one -dimensional statistical conformational model to permit calculation of 
properties of large linear molecules. Each bond is permitted to assume 
several fixed-geometry states. For equilibrium properties, nearest neighbor 
interactions between these states, as incorporated by matrix methods, suffice 
to produce agreement between numerous calculated results and experimental 
data. 1) Even positive moments of intra-chain distances have been calculated 
by methods proposed previously. A general method of expanding the d.f. in 
terms of moments and any approximate d.f. was developed. Also, bounds on 
inverse moments and higher positive moments were foirnulated from known 
moments and the additional constraints of maximum and minimum distances 

117 



accessible to a chain molecule. 2) Non -equilibrium properties are treated 
by assigning rates for transitions between states for each bond. A real- 
istic treatment, including neighbor dependence of these rates, was derived; 
this theory is valid for linear molecules of any size. 

Major Findings: 

1) Previous expansions of distance d.f.'s were unsatisfactory 
because the approximating d.f. generally used was the Gaussian function 
which is inadequate for shorter chains. The present formulation permits 
one to obtain better approximate d.f.'s by using the freely jointed chain 
d.f.'s as the approximate functions. By use of moments, upper and lower 
bounds can be placed upon the mean inverse distance. (This quantity arises 
in theories of hydrodynamic properties.) These bounds offer better approxi- 
mations than were previously available with a slowly converging series. 2) 
The neighbor dependence of barriers to internal rotation was calculated for 
hydrocarbon chains. For calculating time correlation functions, this neigh- 
bor dependence is important and can be included in a sinple matrix formula- 
tion. 

Significance to Biomedical Research: 

These calculations allow a prediction of a variety of physical 
properties for linear molecules of any length and with any given sequence 
of monomer units , whether in biological or synthetic macromolecules . 
Performing such calculations in advance of laboratory experiments can provide 
the experimental scientist a guide to indicate the value of particular 
experiments . 

Honors and Awards: None 

Publications: 

Jemigan, R. L.: Internal Relaxation in Short Oiains Bearing 
Terminal Polar Groi5)s, in Dielectric Properties of Polymers , pp. 99-128, 
Plenum Publishing Corp., New York, 1972. 



118 



Serial No. 5.4 



1. Physical Sciences Laboratory 

2. Not Applicable 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Biophysical Analysis 

Previous Serial Niimber: 5.5 

Principal Investigator: Ralph J. Nossal, Ph.D. 

Other Investigators: None 

Cooperating Iftiits: Harold Lecar, Ph.D., Biophysical Laboratory, NINDB 

Man Years: 



Total : 


0.4 


Professional: 


0.4 


Other: 


0.0 



Project Description: 

Ob j ectives : 

To supply theoretical foundations and experimental models for 
various observations which arise in physiology and biophysical chemistry. 
The following problems have received particular attention during the past 
year. 

1) Models for cell growth and motility: The objective is to 
provide mathematical descriptions of properties of biological cell popula- 
tions. 

2) Fluctuation phenomena in nerves: The objectives are to relate 
the fluctuations in firing thresholds of nerves to the chemical and physical 
processes underlying excitation and to analyze existing data in order to 
test theories concerning mechanisms of transport of ions across nerve 
membranes . 

Methods: 

1) Kinetic equations are derived and solved by enploying advanced 
mathematical methods. 



119 



2) Laboratory techniques o£ microbiology and electrophysiology are 
combined with techniques o£ mathematical physics. 

3) Existing neurophysiological equations (the Hodgkin -Huxley equations) 
have been modified to include fluctuating forces. Methods of non- linear 
mechanics and statistical physics have been applied to analyze the equations 
and digital conputers are losed to determine phase trajectories. 

Maj or Findings : 

1) Expressions have been derived which relate the movement of 
bands of chemotactic bacteria to relevant metabolic and response parameters. 
Experiments have been performed which substantiate the theoretical predictions, 

2) Analytical expressions have been obtained relating axon-firing 
probability to stimulus duration and, also, describing the distribution of 
latency times as a function of stimulus strength. 

Significance to Biomedical Research: 

1) Kinetics of chemotactic bacteria: Chemotaxis seems to be 
implicated in the recognition of bacteria by leukocytes , the migration of 
monocytes to regions of tissue damage, and processes of cell aggregation. 
Bacterial chemotaxis is an analog phenomenon which is more amenable to 
analytic study. 

2) Threshold fluctuations: It is believed that the study will 
facilitate better understanding of the physical processes underlying the 
excitation and propagation of nervous impulses. 

Honors and Awards : None 

Publications : 

Lecar, H. and Nossal, R. J.: Theory of threshold fluctuations in 
Nerves. I. Response of the Hodgkin-Huxley axon to electrical noise, 
Biophys. J. 11, 1048-1067, 1971. 

Lecar, H. and Nossal, R. J.: Theory of threshold fluctuations in 
Nerves. II. Analysis of various sources of membrane noise, Biophys. J. 11, 
1068-1084, 1971. ~ 

Nossal, R. J.: Boundary movement of chemotactic bacterial 
populations, Math. Biosci . 13_, 397-406, 1972. 

Nossal, R. J.: The growth and movement of rings of chemotactic 
bacteria, Exp. Cell. Res , (to appear) . 



120 



Serial No. _ 5.5 

1. Physical Sciences Laboratory 

2. Not Applicable 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 throu^ June 30, 1972 

Project Title: Carbon-13 Nuclear Magnetic Resonance Studies of Peptides 

and Proteins, Including the use of Selective C^^ Enrichment 

Previous Serial Number: 5.6 

Principal Investigator: Jack S. Cohen, Ph.D. 

Other Investigators: I. Chaiken, Ph.D., LCB:NIAMD; M. Freedman, University 
of Toronto, Toronto, Canada; J. Griffin, Ph.D. LCB: 
NIAMD 

Cooperating Units: Laboratory of Chemical Biology, NIAMD 

Man Years: 



Total : 


0.2 


Professional : 


0.2 


Other: 


0.0 



Project Description: 
Objectives: 

1. To apply carbon-13 Fourier transform NMR methods to macro - 
molecules . 

2. To obtain information regarding the structure and function of 
peptides and proteins. 

Methods : 

1. Synthesis of peptides containing selected C enriched amino 
acids . 

1 7 

2. Use of C Fourier Transform NMR to study these resonances 
in the peptides and the fonnation of active enzyme conplexes . 

13 

3. The study of pH and other dependences of C resonances of 

amino acids and peptides. 



121 



Major Findings: 

1. A peptide corresponding to the 1-15 sequence o£ the amino - 
terminal region o£ ribonuclease was synthesized containing 15% C^^ enriched 
phenylalanine at position 8. Its C^^ resonances could be readily seen 
above the background o£ those due to the natural abundance C^^ (II) o£ the 
other 14 amino acids. Other such peptides have been and are being synthe- 
sized. 

2. Conparison o£ spectra o£ the 1-13, 1-15, and 1-20 peptides 
with each other and with amino acid spectra allows the de£initive assign- 
ments o£ most o£ the resonances. Simulation o£ C^^ NMR spectra using the 
MLAB system greatly assists these assignments. 

13 

3. The pH-dependence of C resonances o£ several amino acids 

and peptides is somewhat surprising. Little effect of the carboxyl 
titration was observed, but fairly large effects of the amino terminus 
titration. This is being investigated for possible applications. 

4. Extension of these studies to several proteins at natural 
abundance of C^^ indicates definite differences between the spectra of the 
nature and denatured fomis, but so far it has not been possible to utilize 
these differences for further studies. 

Honors and Awards : None 

Publications : 

Freedman, M. , Cohen, J. S., and Chaiken, I.: Carbon- 13 Fourier 
Transform Nuclear Magnetic Resonance studies of peptides. Biochem. Biophys . 
Res. Comm. , 43, 1148-1155, 1971. 



122 



Serial No. _ 5.6 

1 . Physical Sciences Laboratory 

2. Not Applicable 

3 . Bethesda 

PHS-NIH 

Individual Project Reports 

July 1, 1971 through June 30, 1972 

Project Title: Excitation and Transport Properties of Fluids; Laser 
Scattering 

Previous Serial Number: 5.7 

Principal Investigator: Ralph J. Nossal, Ph.D. 

Other Investigators: Mildred L. NfcNeel 

Cooperating Units: John Utting, Ph.D., LBP, NIAMD; Leonard Kohn, M.D. , 
LPB, NIAMD 

Man Years : 

Total: 0.6 

Professional: 0.6 
Other: 0.0 

Project Description: 

Objectives: 

To proA/lde basic knowledge concerning the excitation properties 
and transport properties of both siirple fluids and conplex solutions con- 
taining biological macromolecules . 

To develop a laser light scattering spectrometer to measure 
hydrodynamic coefficients of biological macromoleciiles , the rate constants 
of bimolecular reactions, and the swimming speed distribution of motile 
microorganisms . 

Nfethods : 

Theoretical techniques of mathematical physics and statistical 
mechanics are enployed in order to develop new physical theories. 
Theoretical studies are performed in support of new experiments, particular- 
ly those involving the scattering of laser light from biological materials. 
A correlation- function laser scattering spectrometer is being developed 
under contract. 



123 



Major Findings: 

Techniques have been developed for using a laser light scattering 
correlation spectrometer for the purpose of measuring the diffusion 
coefficients of biological molecules having a wide range of sizes. The 
theoretical basis of light scattering from motile microorganisms has been 
investigated, and experiments have been performed which establish the 
feasibility of using laser light scattering spectroscopy as a new quantita- 
tive assay for bacterial motility. Methods have been developed for detect- 
ing spatially anisotropic motion, applicable e.g., to studies of chemotactic 
response and measurements of electrophoretic mobilities . 

Significance to Biomedical Research: 

Almost all biological phenomena occur in a fluid environment. A 
number of fundamental questions concerning the physical behavior of fluids 
yet remain unanswered. Their elucidation will ultimately facilitate better 
understanding of the functions and properties of biological systems. 

The development of the laser correlation spectrometer shoiild 
enable rapid and precise measurement of various physical parameters which 
characterize systems of biological molecules and motile microorganisms. 

Honors and Awards : None 

Publications: 

Nossal, R. J., Chen, S-H. and Lai, C. C. : Use of laser scatter- 
ing for quantitative determinations of bacterial motility. Optics Comm . 
4, 35-39, 1971. 

Nossal, R. J. and Chen, S-H.: Light scattering by motile bacteria. 
J. de Physique , (to appear) . 

Nossal, R. J. and Chen, S-H.: Laser measurements of chemotactic 
Response of Bacteria , Optics Comm . (to appear) . 



124 



Serial No. 5.7 

1. Physical Sciences Laboratory 

2. Not Applicable 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Intermolecular Forces in Biological Structures 

Previous Serial Number: 5.8 

Principal Investigators: V. Adrian Parsegian, Ph.D., George H. Weiss, Ph.D, 

and David Gingell, Ph.D. 

Other Investigators: None 

Cooperating Units : None 

Man Years : 



Total : 


2.0 


Professional: 


1.6 


Other : 


0.4 



Project Description: 

Objectives: 

To identify and learn to calculate those intermolecular forces 
governing the structure of matter at the biological level . These have been 
Coulombic (electrostatic) interactions between charged species and van der 
Waals (electrodynamic) forces. 

Methods: 

Classical and quantum-mechanical treatment of electromagnetic 
and statistical -mechanical behavior as well as molecular models of specific 
structural interactions. 

Major Findings: 

Long-range physical forces are strong enough to hold biological 
cells together. These forces may be sufficiently specific between cells 
of like tissue type to account for the fundamental observation that cells 
will sort themselves out of a mixture to form aggregates of only one tissue 
type. 



125 



In addition to the use of measured absorption spectra to conpute 
electrodynamic forces , we have learned to use data derived from reflection 
spectra to make more reliable estimates. 

The theory of van der Waals interactions has been developed to give 

the interaction between anisotropic and inhomogeneous bodies . Also there 

is now a formulation of the interaction between long thin molecules in 
solution. 

Significance to Biomedical Research: 

The physical theory of intercellular forces provides a strong 
logic for understanding and controlling cell -cell interaction. In particular 
it is leading to experiments where the outer cell surface may be altered by 
binding of known substances which will in turn promote aggregation of cells. 
We are trying to steer this effort to convert non-aggregative cell popula- 
tions such as metastasizing cancer cells into cellular aggregates sticking 
either to themselves or better to an inserted artificial substrate. 

Honors and Awards: None 

Publications: 

Ninham, B, W. , Parsegian, V. A. and Weiss, G. H.: On the macro- 
scopic theory of tenperature -dependent van der Waals forces. Journal of 
Statistical Physics 2, 323-328, 1970. 

Parsegian, V. A. and Weiss, G. H.: Dielectric anisotropy and the 
van der Waals interaction between bulk media. J. Adhesion 3,259-267, 1972. 

Parsegian, V. A. and Weiss, G. H.: On van der Waals interactions 
between macroscopic bodies having inhomogeneous dielectric susceptibilities. 
J. Colloid and Interface Sci . (in press), 1972. 

Parsegian, V. A.: The non-retarded van der Waals interaction 
between thin rods at all angles. J. Chem. Ph^^s . (in press), 1972. 

Parsegian, V. A. and Gingell, D. : A physical force model of 
biological membrane interaction, in Recent Advances in Adhesion ed. 
L. H. Lee, Gordon and Breach Pub. Co., New York and London (in press), 1972. 

Parsegian, V. A. and Gingell, D. : Some features of physical 
forces between biological cell membranes. J. Adhesion (in press), 1972. 

Parsegian, V. A. and Gingell, D. : On the electrostatic inter- 
action across a salt solution between two bodies bearing unequal charges. 
Biophysical Journal (in press), 1972. 

Gingell, D. and Parsegian, V. A.: Computation of van der Waals 
interactions in aqueous systems using reflectivity data. J. Theor. Biol , 
(in press) , 1972. 

126 



Parsegian, V. A. and Ninham, B. W. : van der Waals forces in 
many-layered structures: Generalizations of the Lifshitz result for two 
semi -infinite media. J. Theor. Biol, (in press) , 1972. 



127 



Serial No. 5.8 

1. Physical Sciences Laboratory 

2. Not Applicable 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Consulting Services 

Previous Serial Number: 5.9 

Principal Investigators: Richard I. Shrager, George H. Weiss, Ph.D. 

Other Investigators: Mildred McNeel, James A. Ferretti, Ph.D., Adrian 
Parsegian, Ph.D. 

Cooperating Units: Elliot Chamey, Ph.D., Ryo Hirasawa, Ph.D., William 
Eaton, Ph.D., Edwin Tucker, Ph.D., Norman Sharpies s, 
Ph.D., Laboratory of physical Biology, NIAMD; Elemer 
Mihali, Laboratory o£ Biochemistry, NIAMD; Eitan Rotlevi, 
M.D. , Florence Millar, M.D., Laboratory o£ Physiology, 
NCI; Robert Highet, Ph.D., Lipid Metabolism Branch, 
NHLI; William Caveness, M.D. , Laboratory o£ Experimental 
Neurology, NINDS; Eugene Fischmann, M.D. , Freedman's 
Hospital; David Rodbard, M.D., Reproduction Research 
Branch, NICHD. 

Man Years: 

Total : 1.7 

Professional: 1.5 
Other: 0.2 

Project Description: 

Objectives: 

To provide consulting services in applied mathematics, biometry, 
theoretical chemistry, theoretical physics, and various aspects of numerical 
analysis to workers primarily in experimental fields. 

Methods : 

A considerable amount of consulting is done in the application of 
curve fitting techniques to the results of biological and biochemical experi- 
ments. Other consulting involves applications of statistics to medical data 
and the application of a program developed by Dr. Ferretti of this Laboratory 
for the determination of physically interesting parameters from NMR data. 



128 



Major Findings: 

Long term effects of head injuries in the Korean War have been examined 
by studying work status from a Red Cross survey taken approximately seven- 
teen years after injury. So far it has been found that a certain fraction 
of the head injured did not work at the time of the survey. This fraction 
was larger than that for a conparable set of controls. However, no feature 
of the injury predominated in indicating who would or would not work. 
Further, no differences appeared in the quality of work between the head 
injured and controls. 

A minimal set of two criteria for vectorcardiography was developed, 
from a study of observer variation from a set of thirteen criteria tested 
with four readers. 

Analytical solutions were found to the equations describing the 
immunoradioassay technique, and these have been applied to several common 
experimental procedures . 

Honors and Awards : None 

Ptiblications : 

McNeel, M. L., Becker, E. D. , Shoip, R. R. : An evaluation of the 
nuclear magnetic resonance total line shape of the uncoupled, exchanging 
two -site systems. Journal of Physical Chemistry , 76 , 71-78, 1972. 

Weiss, G. H. , Barber, M. R. , Fischmann, E. J.: Multielectrode grids 
which measure torso area and resistivity and yield dipole moment calibrated 
for these variables. Proceedings of the Xlth International Vectorcardio - 
graphy Syiiposium , 30-41, 1971. 

Weiss, G. H., Fischmann, E. J., Barber, M. R. : In vivo determination 
of orthogonal lead vectors . Proceedings of the SateTTite Symposium of the 
XXVth International Congress of Physiological Sciences (to appear) . 

Weiss, G. H., Caveness , W. : Prognostic factors in the persistence of 
posttraumatic epilepsy. Jovimal of Neurosurgery (to appear) . 



129 



Serial No. 5.9 

1 . Physical Sciences Laboratory 

2 . Not Applicable 

3. Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: Fundamental Studies 

Previous Serial Number: 5.10 

Principal Investigators: George H. Weiss, Ph.D., Richard I. Shrager, 

Andrew G. De Rocco, Ph.D. 

Other Investigators: James E. Kiefer, Mildred McNeel 

Cooperating Units: Milton Sobel, Ph.D., University of Minnesota; 

Menachem Dishon, Ph.D., Weizmann Institute o£ Science, 
Israel 

Man Years : 

Total: 2.0 

Professional: 1.8 
Other: 0.2 

Project Description: 

Objectives: 

This project involves several miscellaneous areas of investigation. 
These include : 

1. Development of the theory of comparative clinical trials so as to 
maintain a given level of accuracy while subjecting as few people as possible 
to the poorer of two (or more) treatments . 

2. Development of numerical algorithms for the solution of badly be- 
haved differential equations. 

3. Development of approximate solutions to difference differential 
equations that are used in the quantitative analysis of ecological problems 
and in the theory of epidemics. 



130 



Methods : 

A variety o£ mathematical methods are used. These include 
extensions of the predictor-corrector method in the numerical solution o£ 
differential equations, methods of classical analysis related to statistics, 
and perturbation theory. 

Major Findings: 

A variable step technique for the numerical solution of differ- 
ential equations has been developed. This allows the program to choose 
the step size adaptive ly, making the method particularly suitable for the 
equations of enzyme kinetics in which rate constants may be greatly different. 

In earlier work on conparative clinical trials we have found that 
it is possible to reduce the number of patients assigned to the poorer of 
two treatments provided that the success probabilities are known a priori . 
Usually this is not the case. We have found that a short preliminary trial 
to get a rough estimate of success probabilities does not degrade the per- 
formance of the trial designs considered earlier, by a significant amount. 
Further, we have considered the case in which the results of a treatment 
are not immediately observable but may be delayed by a random amount of time. 
This problem was analyzed using a likelihood procedure and a considerable 
savings in the number of people assigned to the poorer treatment was found 
feasible by appropriate sampling methods. 

Honors and Awards : None 

Publications: 

Shrager, R.I.: Quadratic programming for nonlinear regression. 
Communications of the Association for Computing Machinery 15 , 41-45, 1972. 

Weiss, G. H. and Sobel, M. : Play-the-wixmer rule and inverse 
sanpling in selecting the better of two binomial populations. Journal of 
the American Statistical Association 66 , 545-551, 1971. 

Weiss, G. H. and Sobel, M. : Some results on using the play-the- 
winner rule with binomial problems. Proceedings of the Sixth Berkeley 
Synposium (to appear) . 

Weiss, G. H. and Sobel, M. : A conparison of pi ay -the -winner and 
vector-at-a-time sampling for selecting the better of two binomial popula- 
tions with restricted parameter values. Trabajos de Estadistica y de 
Investigacion Operativa XXII , 195-206, 1971. 

Weiss, G. H. and Blumenfeld, D. E. : Merging from an acceleration 
lane. Transportation Science 5, 161-168, 1971. 



131 



Weiss, G. H. and Dishon, M.: On the asynptotic behavior o£ the 
stochastic and deterministic models o£ an epidemic. Mathematical Biosciences 
11, 261-265, 1971. 

Kiefer, J. E. and Weiss, G. H. : A truncated test for choosing the 
better o£ two binomial populations. Journal of the American Statistical 
Association 66, 867-871, 1971. 

Weiss, G. H. and Dishon, M. : Some economic problems related to bum- 
in programs. IEEE Transactions on Reliability R-20 , 190-194, 1971. 

Weiss, G. H. and Dishon, M. : Two problems in the maintenance of 
communications satellites. Israel Journal of Technology 9, 613-616, 1971. 

Weiss, G. H.: Some approximate solutions to the master equation. 
Journal of Statistical Physics (to appear) . 

Weiss, G. H. and Sobel, M. : PI ay -the -winner rule and inverse sanpling 
for selecting the best of k ^ 5 binomial populations. Annals of Mathe - 
matical Statistics (to appear) . 

Weiss, G. H. : On a perturbation method for the theory of epidemics. 
Advances in Applied Probability 5, 221-222, 1971. 

Weiss, G. H., Hoel, D. G. , and Sobel, M. : A two stage procediore for 
ciioosing tl>e better of two binomial populations. Biometrika , V. 59_, 1972. 

Wulf, A. and De Rocco, A. G.: Statistical mechanics for long semi- 
flexible molecules: A model for the nematic mesophase. J. Chem. Phys . 
55_, 1, 12-27, 1971. 

York, E. D. and De Rocco, A. G.: Two-body dispersion forces for pairs 
of polyatomic molecules. J. Chem. Phys . 55_, 4, 1900-1903, 1971. 



132 



Serial No. 5.10 



1 . Phys i cal Sciences Laboratory 

2. Not Applicable 

3 . Bethesda 

PHS-NIH 
Individual Project Reports 
July 1, 1971 through June 30, 1972 

Project Title: Nuclear Magnetic Resonance (NMR) Studies o£ Proteins 

Previous Serial Number: 5.11 

Principal Investigator: Jack S. Cohen, Ph.D. 

Other Investigators: A. Schechter, M.D. , J. Griffin, Ph.D., Laboratory of 
Chemical Biology, NIAMD; H. Yeh, Ph.D., L. Cohen, 
Laboratory of Chemistry, NIAMD; M. McNeel, R. I. 
Shrager; P. Cohen, Ph.D. C.E.N.S. (France) and 
R. Feldman, PDP-10, DCRT 

Cooperating Units: Laboratory of Chemical Biology, NIAMD; Laboratory of 
Chemistry, NIAMD 

Man Years: 

Total: 1.0 

Professional: 0.8 
Other: 0.2 

Project Description: 

Objectives: 

1. To obtain detailed information on the structure and function of 
enzymes, both qualitatively in terms of the groups involved and quantitative- 
ly in terms of the molecular interactions. 

2. To utilize the histidine NMR titration curves as a sensitive 
monitor of protein interactions at the molecular level. It is noted that 
histidine is the only amino acid with a pK around neutrality which gives 
rise to well resolved titrating resonances. 

3. To extend the applicability of high resolution NMR as a quanti- 
tative analytical tool to proteins. 

Methods : 

The region of the NMR spectrum containing the histidine imidazole 
C2-H resonances of a number of proteins is subjected to detailed analysis. 
The 220 MHz time -averaged spectra are digitized and fitted with a series 
of Lorentzian curves using the MODELAIDE least squares fitting program 



133 



with the on-line IBM 2250 display unit. The chemical shifts o£ the 
imidazole resonances as a function of pH are fitted to several mechanistic 
models using the MLAB system of the DEC PDPIO conputer. Difference spectra 
between different states are obtained utilizing a con5)uter program. 

Major Findings: 

1. The acid denaturation process of Staphylococcal nuclease was 
followed by measuring the areas of the peaks. The results indicated that 
the renaturation process is reversible and that it occurs via multiple 
equilibria with different nucleation sites. 

2. A study of bovine and human carbonic anhydrases has indicated 
that no histidine group is directly involved in their enzymatic mechanism. 
However, a conformational change on binding inhibitor occurs in HCA, 
resulting in the observation of an extra histidine titration -curve with a 
pK of 6.04. 

3. Titration curves of the proton resonances of imidazole in 
histidine, its blocked derivatives and small peptides show the effects of 
the adjacent titrating carboxyl and amino groups. Mathematical models 
have been developed which describe these curves and enable the estimation 
of the microscopic ionization constants . 

4. Extensive data on the titration curves of the two active site 
histidines (at positions 12 and 119) in ribonuclease has been obtained. 
When the equation for two interacting titrating groips is applied to the 
asymmetric curves of the active site histidines it is found that, contrary 
to some interpretations , the two histidines are not mutually interacting, 
but are each interacting with other groijps in the protein. 

5 . The nature of this interaction is greatly affected by the 
addition of conpetitive inhibitors which bind in the active site. We are 
currently analyzing titration data for the C2-H resonances resulting 
from phosphate , sulphate , mononucleotides and a dinucleotide phosphate 
substrate analog (sipplied by G. Jones of Syntex, Inc.) . Our results 
indicate that the previous assignments of the histidine resonances is 
incorrect, and that contrary to all previous mechanisms proposed for ribo- 
nuclease, the active site histidines are not protonated at pH values where 
the enzyme is most active (pH 7-8) . 

6. A series of model conpounds are being studied as a means of 
obtaining data with which to interpret observed effects on imidazole 
residues in proteins. These include a series containing imidazole separated 
from a carboxyl group by one , two and three methylene groups , and which are 
locked in a single conformation (prepared by Dr. Kikugawa in Dr. Louis 
Cohen's laboratory). A series of fluorine -containing conpounds are also 
under study. So far the ^^F as well as the proton NMR titration curves 
have been measured for 2- and 4- fluoro- imidazole and 4 -fluoro -histidine 
(prepared by Dr. Kirk in Dr. Louis Cohen's laboratory). These show some 



134 



very interesting results, and electron density calculations are being 
made in an attenpt to understand the opposite shift effects noted for the 
2- and 4-fluoro imidazoles on increasing pH. 

7. Studies with sperm whale and horse myoglobins indicate seven 
titrating histidines (out of 12) in the former case and six (out of 11) 
in the latter. By consideration of the differences between these two 
proteins one of the titrating resonances in sperm whale myoglobin can be 
selectively assigned to His -12. One titration curve in each case has a 
much higher pK value than normal (8.05 in sperm \^^ale myoglobin) and is 
shifted to higher field values. The possible reasons for this effect and 
the assignment of this curve in terms of our knowledge of the X-ray 
structure, and the effects of ligands, is being considered. Preliminary 
results indicate that this curve may be a monitor of activity of the 
protein. 

8. Similar studies are being conducted on cytochrome-c. Only 
one titrating histidine resonance is seen in this case out of the three 
histidines present. 

9. The interaction of a peptide hormone, oxytocin, with neuro- 
hypophyseal proteins has been studied. The resLilts indicate that the 
tyrosine residue on the peptide is intimately involved in the binding 
process . 

10. A conputer-based molecular display system has been set up (in 
collaboration with R. Feldmann) in order to inprove techniques for the 
observation of groups within a protein molecule whose X-ray structure is 
known. The system is also being developed for the manipulations of groi:ps 
within a macromolecule, in connection with energy minimization calculations. 
It is hoped to apply this specifically to ribonuclease- inhibitor inter- 
actions for the calculation of the conformation of enzyme -inhibitor 
complexes. 

Significance to Biomedical Research: 

1. A general and objective method for the analysis of resolved 
resonances in protein IMMR spectra has been developed. These methods have 
shed light on the mode of action of S. nuclease, carbonic anhydrases and 
other proteins. 

2. A recently conpleted study of the acid denaturation of 
nuclease provides definite evidence for a multiple step, as opposed to a 
two -state, equilibrium during denaturation, a point in contention for a 
long time between protein biochemists. 

3. An analysis of the titration data for pancreatic ribonuclease 
is under way. This approach provides in principle a means of distinguish- 
ing between published mechanisms of its activity, and obtaining a detailed 
analysis of its mode of action. 



135 



4. Results for myoglobin indicate a unique means for the analysis 
of structure -function relations, which should be extendable to haemoglobin. 

Honors and Awards: None 

Publications: 

Cohen, J, S. , Fell, M. , and Chaiken, I.: Proton Magnetic 
Resonance Studies of the tyrosine residues of Staphylococcal nuclease 
using [3,5- H2] tyrosine. Biochem. Biophys. Acta ., 236 , 468-478, 1971. 

Epstein, M. F., Schechter, A. N. and Cohen, J. S.: Folding of 
Staphyloccal nuclease : Magnetic resonance and fluorescence studies of 
individual residues. Proc. Nat. Acad. Sci. (U.S.) , 68, 2042-2046, 1971. 

Sachs, D. H. , Schechter, A. N. and Cohen, J. S.: Nuclear magnetic 
resonance titration curves of histidine ring protons. J. Biol. Chem. 
246, 6576-6580, 1971. 

Cohen, J. S. , Yim, C. T., Kandel, M. , Gomall, A. C, Kandel, S. J 
and Freedman, M, H. : Studies of the histidine residues of carbonic an- 
hydrases using high-field proton magnetic resonance. Bioche mistry , 11, 
327-334, 1972. 

Shrager, R. I., Cohen, J. S., Heller, S. R., Sachs, D. H. , and 
Schechter, A. N. : ^fa.thematical models for interacting groips in nuclear 
magnetic resonance titration curves. Biochemistry, 11, 541-547, 1972. 



136 



Serial No. 5.11 



1. Physical Sciences Laboratory 

2. Not Applicable 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1971 through June 30, 1972 

Project Title: High-Resolution Carbon-13 Fourier Transform NMR 
Spectroscopy at 55 MHz 

Previous Serial Number: 5.12 

Principal Investigator: James A. Ferretti 

Other Investigators: Edwin D. Becker, Ph.D., Laboratory of Physical 

Biology, NIAMD; Victor Colbum and Donald Deutsche, 
Computer Systems Laboratory, DCRT; Thomas Clem, 
Bioelectrical Engineering DRS 

Cooperating Units: Laboratory of Physical Biology, NIAMD and Bio- 
electrical Engineering, DRS; Computer Systems 
Laboratory, DCRT 

Man Years: 

Total : 0.8 

Professional: 0.8 
Other: 0.0 

Project Description: 

Objectives: 

To develop equipment and methods to obtain natural abundance 
carbon-13 NMR spectra at 55 MHz (ca. 51 kgauss) using the Fourier trans- 
form technique. 

Methods : 

Equipment is being constructed, preliminary tests are being 
performed, programs are being written, and a Raytheon 704 coirputer has 
been interfaced to the spectrometer to permit recording the free induction 
decay in a pulse NMR experiment. The Fourier transform after sufficient 
signal enhancement is then taken. 

Significance to Biomedical Research: 

Carbon-13 NMR provides a sensitive technique to correlate 



137 



i«3lecular changes with biological activity such as changes in confoimtion 
and active binding sites in enzymes. 

Honors and Awards: None 

Publications : None 



138 



July 1, 19 71 through June 30, 19 72 



PUBLIC HEALTH SERVICE - NATIONAL INSTITUTES OF HEALTH 
DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 



Summary of Branch Activities 



2. HEURISTICS LABORATORY 



DCRT 6 



Serial Number 

3. James R. Slagle 
Chief 



The Heuristics Laboratory continued at a high rate of purposeful activity this 
year as one of the major centers investigating the frontiers of computer 
applications, compiling a remarkable record for a comparatively recently 
established group. 

For example, three papers from our group appeared at the Second International 
Conference on Artificial Intelligence. Also a new book "Symbolic Logic and 
Mechanical Theorem Proving" by Chin-Liang Chang and Richard C. T. Lee, was 
delivered to the publisher, while James R. Slagle's "Artificial Intelligence: 
The Heuristic Programming Approach" went into a second printing. 

The ongoing projects of question answering (DCRT 6.1) and pattern recognition 
(DCRT 6.3) are continuing to bear fruit in interesting ways. For example, 
the question answering project has led by three separate paths to advances in 
the fashionable area of theoretical applications to programming now 
established as a new project (DCRT 6.6). This includes the work on partial 
evaluation of programs, secondly work on program analysis and synthesis by 
theorem proving techniques and third, applying some of our specially 
developed methods to problems arising in proving properties of specific 
programs. Also, that sub-section of project DCRT 6.1 dealing with linear 
inequalities has led to an application in optimization of external beam 
radiation treatment planning. The pattern recognition project has developed 
new cluster analysis techniques which hold great promise in biology. 

The radiation therapy and cluster analysis projects point to another 
significiant trend in our work. This is the use of semiautomatic or 
interactive systems. The trend first began on a large scale in our group 
with the on-line chemical information work (DCRT 6.4). This work has made great 
strides this year with the addition of a chemist to our staff (see DCRT 6.90- 
6.93). The showcase of interactive computing is of course the MODELAB project 
DCRT 6.5, which has received widespread acceptance during the past year. It 
allows the user powerful facilities in fitting models to his data. Here we 
begin to get the essence of our idea of interactive computing. The machine 
becomes a capable partner rather than just an efficient servant. The cluster 
analysis and radiation therapy planning programs are interactive in that sense. 
In cluster analysis (DCRT 6.3), the points are mapped by the computer into 
two dimensions and the user visually decides which are the clusters. In the 
semiautomatic optimization of external beam therapy planning (DCRT 6.7) 
the user decides on a good plan by looking at an isodose pattern after the 

139 



computer does some iterations of optimization. In both cases the philosophy 
is to have the computer and the user do what each does best. 

There were also some new developments in the information retrieval (DCRT 6.2) 
and the systems support projects (included in DCRT 6.6), but these are now 
out of the mainstream of our group. We have also provided some high level 
consulting and extensive assistance (DCRT 6,80-1), to selected projects in 
DCRT and elsewhere. One of these projects, intensive care monitoring, 
Computer Systems Laboratory (DCRT 4.1]) has a long term commitment from one 
member of our group. 



140 



Serial No. DCRT 6.1 

1. Heuristics Laboratory 

2. 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 19 72 

Project Title: Automatic Question-Answering and Problem Solving 

Previous Serial Number: Same 

James R. Slagle, Ph.D. 



Principal Investigator: 
Other Investigators: 

Cooperating Units: 
Man Years: 



Chin-Liang Chang, Ph.D., John K. Dixon, Ph.D., 
Louis Modes, Ph.D., Thomas L. Jones, Ph.D., 
Richard C. T. Lee, Ph.D., Lewis M. Norton, Ph.D. 

None 



Total: 


4.0 


Professional: 


3.3 


Other: 


0.7 



Project Description: 

Objectives: 

It is widely recognized that advanced question answering systems will 
need to perform logical deduction. Thus, one aspect of our work is the 
creation of efficient inference systems and strategies. Another is 
the study of goal-structured problem solving, where general criteria can 
be developed for which subgoal to attempt next. A third aspect is 
linguistics and the formalization of verbal data. 

Methods Employed: 

Theoretical studies in symbolic logic and combinatorics to develop 
heuristics that increase efficiency. Also computer experiments of new 
strategies . 

Major Findings: 

(1) Further work was done on "building in" valuable concepts such as 
commutativity and associativity into resolution systems. These concepts 
occur in such basic theories as number theory and set theory. 



141 



Commutativity requires a fairly straightforward modification to the 
unification (matching) algorithm. Associativity and (left and right) 
identity are more complicated. 

(2) It was shown that a statement can be proven by generating a closed 
pseudo-semantic tree. 

(3) A method was devised to test the backing-up densities idea for 
deciding on the existing MULTIPLE program. 

(4) A study was performed on the art of using predicate calculus. 

Significance to Biomedical Research: 

Sophisticated question answering and problem solving methods will 
eventually be applied to biomedical research. See Projects 6.6 and 6.7 
which are offshoots of this project. 

Proposed Course: 

While a continuing effort is maintained on the most promising aspects 

of this work, much of the manpower is being channelled into more immediately 

applicable areas, making use of some techniques developed here. 

Publications: 

1. Chang, C.L. and Lee, R.C.T.: Symbolic Logic and Mechanical Theorem 
Proving , New York Academic Press; not yet published. 

2. Chang, C.L. and Slagle, J.R. : An Admissible and Optimal Algorithm for 
searching AND/OR graphs. Artificial Intelligence , 2: 117-128, 1971. 

3. Dixon, J.K.: Z-Resolution proving theorems with compiled axioms. 
J. Assoc. Comm. Mach. (to be published) 

4. Hodes, L.: Solving problems by formula manipulation in logic and 
linear inequalities. Artificial Intelligence , to be published. 

5. Lee, R.C.T.: Fuzzy logic and the resolution principle. 
J. Assoc. Comp. Mach. 19: 109-119, 1972. 

6. Lee, R.C.T. and Chang, C.L.: Some properties of fuzzy logic. 
Information and Control, 19: 417-431, 1971. 

7. Norton, L.M.: Experiments with a heuristic theorem-proving program 
for predicate calculus with equality. Artificial Intelligence , 

2, 261-284, 1971. 

8. Simon, R. and Lee, R.C.T. : On the optimal solution of AND/OR series- 
parallel graphs. J. Assoc. Comp. Mach. 18: 354-372, 1971. 



142 



J. Assoc. Comp. Mach. . 19: 120-135, 1971. 

8 i- in Awu/UK trees. Information Sciences . 3: 315-342, 1971. 



143 



Serial Number DCRT 6.2 

1. Heuristics Laboratory 

2. 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Storage and retrieval studies 

Previous Serial Number: Same 

Principal Investigator: Gary D. Knott 

Other Investigators: None 

Cooperating Units: None 



Man Years : 






Total: 


0. 


,4 


Professional: 


0. 


,25 


Other: 


0. 


,15 



Project Description: 
Objectives: 

Comprehensive analysis of storage and retrieval algorithms 
Methods Employed: 

Finite algebra number theory and other advanced mathematics 
Major Findings: 

An extensive survey of hash coding methods has been prepared 

Significance to Biomedical Research: 

Hash coding is important in many large storage and retrieval systems 
e.g. the cataloging of bacteria. 

Proposed Course: 

Continued development and wide dissemination of the large amount of 
discovered information. 



144 



Publications; 



Tollutls iTZf'AirZYsZriiT''' ""^ "°"^= -- -"'-»i. 



145 



Serial Number DCRT 6.3 

1. Heuristics Laboratory 

2, 

3. Bethesda 

PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Automatic Pattern Recognition 

Previous Serial Number: Same 

Principal Investigator: James R. Slagle, Ph.D. 

Other Investigators: Chin-Liang Chang, Ph.D.; John Dixon, Ph.D.; 
Louis Hodes, Ph.D.; Thomas L. Jones, Ph.D.; 
Richard C. T. Lee, Ph.D. 

Cooperating Units: CSL-DCRT and NIAMD 

Man Years 



Total: 


2, 


.35 


Professional: 


2. 


.05 


Other: 


0. 


.3 



Project Description: 

Objectives: 

To advance the state of the art of those areas of pattern recognition which 
hold promise for biomedical application. For example, cluster analysis. 

Methods Employed: 

The development and programming of new algorithms to classify patterns 
and aid in cluster analysis. The application of artificial intelligence 
techniques to pattern recognition. 

Major Findings: 

A new algorithm for mapping points into two-dimensions for subsequent 
cluster analysis was developed. This algorithm was programmed and tested 
on a variety of sets of data. It uses a relaxation method and seems quite 
efficient. It was used to cluster data from hand x-rays from a study on 
arthritis . 

A new algorithm for pattern classification by multilayered linear machines 
was developed, ^^7hich should be more efficient. 



146 



Several other tentative studies were undertaken. They included investigation 
of speech synthesis, a start on an automated chemical diagram reader, a 
growth model for embryology. 

Also, one member served on an ad hoc coinmittee to investigate the 
automation of pap smear testing, another on a nuclear medicine project 
which will include automatic image enhancement and processing. 

Significance to Biomedical Research: 

The automatic pattern recognition and classification techniques continue 
to hold great promise to biomedicine. Almost every area has its tedious 
visual analyses and imprecise drawing of conclusions from masses of data. 
Some fields are more ripe for automation than others. 

Proposed Course: 

An intensive effort to uncover more fields that are ripe for any of our 
techniques of cluster analysis or image processing will be undertaken. 

Publications : 

Jones, T.L.: MULTIPAT: A proposed scene analysis language and interpreter. 
Proceedings of the Third Houston Conference on Computers and System Science. 

Jones, T.L.: A computer model of simple forms of learning in infants. 
1972 Spring Joint Computer Conference Proceedings. 

Lee, R.C.T.: Application of information theory to select relevant 
variables. Mathematical Biosciences 11: 153-161, 1971 



147 



Serial Number DCRT 6.4 

1. Heuristics Laboratory 

2. 

3. Bethesda 

PHS - NIH 
Individual Project Report 
July 1, 19 71 through June 30, 1972 

Project Title: Chemical Notation and On-Line Information Systems 

Previous Serial Number: Same 

Principal Investigators: Stephen R. Heller, Deena A. Koniver 

Other Investigators: George Miller, Richard J. Feldmann, CCB. 

Cooperating Units: Computer Center Branch i 

Man Years: 



Total: 


1.0 


Professional: 


.9 


Other: 


.1 



Project Description: 

Objective: 

The automatic encoding and decoding of organic chemical compounds in 
Wiswesser Line Notation (WLN) , a widely used linear notation. The overall 
objective of this project is to enable a chemist to easily communicate with 
computer-based chemical information systems. 

Methods Employed: 

There are two basic I'JLN computer programs being developed for the PDP-IO 
computer. One accepts as input the atom-by-atom connection table of a 
compound and generates the corresponding WLN. The input connection table 
can be obtained by several possible means, including a chemist (or user) 
drawing the structural diagram on the Rand Tablet/display, or by using a 
normal computer terminal. 

The other basic program converts an input string of WLN to its 
corresponding connection table. Since a connection table itself is 
rather unintelligible, a program is then used to generate a display of the 
structural diagram from the connection table. 



148 



Major Findings: 

It is possible to write computer programs to both encode and decode 
compounds as complex as polycyclic fused ring systems, a chain of two 
such ring systems, peri-fused rings, salts, ions, some metallocenes and 
some chelates. 

The WLN project has reached a plateau after being programmed to handle 
salts, ions, ferrocene and chelate structures, and polyfused and 
perifused ring compounds. The program was used to generate WLN for 
1,000 out of 1,500 of the compounds in the second edition of "Psychotropic 
Drugs and Related Compounds" by E. Usdin and D. H. Efron of NIMH. Also 
the WLN for the various NMR catalogs of 742 compounds were generated for 
a search program. 

Proposed Course: 

The feasibility of extending either or both of the programs to handle 
more complex situations, such as bridges and spiro rings, will be studied. 
Unless an actual user of the programs has a real need for the handling of 
such complexities, it is unlikely that the programs will be expanded much 
beyond their present capabilities. 

Significance: 

This project has helped promote a standard notation, facilitating 
communication among chemists. 

Publications: 

Feldmann, R.J. and Koniver, D.A. : Interactive searching of chemical files 
and structural diagram generation from Wiswesser Line Notation, 
J. Chem. Doc. 11: 154-159, 1971. 

Heller, S.R. and Koniver, D.A. : Computer Generation of Wiswesser Line 
Notation, II. Polyfused, Perifused and Chained Ring Systems. 
J. Chem. Doc. 12: 55-59, 1972 

Miller, G.A.: Encoding and Decoding WLN. J. Chem. Doc. 12: 60-67, 1972. 

Koniver, D.A. , Wiswesser, W.J. and Usdin, E.: Wiswesser Line Notation: 
Simplified techniques for converting chemical structures to WLN. 
Science (in press) . 



149 



Serial Number DCRT 6.5 

1. Heuristics Laboratory 

2. 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 19 72 

Project Title: Modelab 

Previous Serial Number: Supersedes 6.5 

Principal Investigator: Gary D. Knott _^ -_._ 

Other Investigators: Douglas Reece, Richard Shrager (PSL) 

Cooperating Units: PSL . ,■ 



Man Years: 






Total: 


1, 


.75 


Professional: 


1. 


.55 


Other 




.2 



Project Description: 

Objectives: 

To allow a scientist user powerful facilities to fit models to his data. 

Methods Employed: 

Curve-fitting, differential equation solver, other elaborate programming 
techniques. The investigator works interactively with the system, 
"conversing" back and forth in a simple easy-to-learn language. He 
never has to do any programming himself. 

Major Findings: 

Over fifty people have used the system. For example, it has been used 
successfully by chemists studying equilibrium phenomena and bond 
interactivity. Physicians have employed it for studying absorption of 
radioactive tracers, neural transfer mechanisms and drug interactions. 

Significance to Biomedical Research : 

It is generally difficult for a biologist to take advantage of what computers 
have to offer. The MODELAB program allows the computer to meet the 
scientist more than halfway. 



150 



Proposed Course: 

The programs are to be further developed and more users actively sought, 



Publications : 



Knott, G.D. and Shrager, R. : On-line modelling by curve fitting. 
Proceedings of SIGGRAPH Symposium on Graphics in Biomedicine, 1972. 



151 



Serial Number DCRT 6.6 

1. Heuristics Laboratory 

2. 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 19 72 

Project Title: Theoretical Applications to Programming 

Previous Serial Number: None (related to 6.1) 

Principal Investigator: James R. Slagle, Ph.D. 

Other Investigators: 



Chin-Liang Chang, Ph.D.; John K. Dixon, Ph.D.; 
Thomas L. Jones, Ph.D.; Louis Hodes , Ph.D.; 
Richard C. T. Lee, Ph.D.; Lewis M. Norton, Ph.D. 



Man Years: 




Total: 


1.3 


Professional: 


1.0 


Other: 


.3 



Project Description: 

Objectives: 

Advancing the understanding of programming and the accuracy and efficiency 
of computer software, automatic programming and program verification. 

Methods Employed: 

The method of partial evaluation of programs (the specializer) was invented 
here (independently at a few other places). Other methods include 
symbolic logic, especially mechanical theorem proving, and heuristic 
approaches. 

Major Findings: 

In the partial evaluation programs heuristics were developed to balance the 
opposing goals of fast execution and economy of storage space. 

The program synthesis methods by theorem proving were combined with partial 
evaluation so that more efficient methods could be produced. 

The assertions stating the corrections of a program were treated as 
examples for theorem proving methods developed under Project 6.1 



152 



Maj or Findings : 

It is possible to write computer programs to both encode and decode com- 
pounds as complex as polycyclic fused ring systems, a chain of two such 
ring systems, peri-fused rings, salts, ions, some metallocenes and 
some chelates. 

The WLN project has reached a plateau after being programmed to handle 
salts, ions, ferrocene and chelate structures, and polyfused and 
perifused ring compounds. The program was used to generate WLN for 
1,000 out of 1,500 of the compounds in the second edition of "Psychotropic 
Drugs and Related Compounds" by E. Usdin and D. H. Efron of NIMH. 
Also, the WLN for the various NMR catalogs of 742 compounds were 
generated for a search program. 

The feasibility of extending either or both of the programs to handle 
more complex situations, such as bridges and spiro rings, will be studied. 
Unless an actual user of the programs has a real need for the handling of 
such complexities, it is unlikely that the programs will be expanded much 
beyond their present capabilities. 

Publications : 

Feldmann, R.J. and Koniver, D.A. : Interactive searching of chemical files 
and structural diagram generation from Wiswesser Line Notation, 
J. Chem. Doc. 11: 154-159, 1971. 

Heller, S.R. and Koniver, D.A. : Computer Generation of Wiswesser Line 
Notation, II. Polyfused, Perifused and Chained Ring Systems. 
J. Chem. Doc. 12: 55-59, 1972. 

Miller, G.A. : Encoding and Decoding WLN. J. Chem. Doc. 12: 60-67, 1972. 

Koniver, D.A., Wiswesser, W.J. and Usdin, E. : Wiswesser Line Notation: 
Simplified techniques for converting chemical structures to WLN. 
Science , (in press) 



153 



Serial Number DCRT 6.7 

1. Heuristics Laboratory 

2. 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 



Project Title: Radiation Treatment Planning 
Previous Serial Number: None 
Principal Investigator: Louis Modes, Ph.D. 
Other Investigators: Frederick L. Faw, NCI. 
Cooperating Units: Radiation Branch, NCI. 
Man Years: 



Total: 


0. 


,52 


Professional: 


0, 


.5 


Other: 


0, 


.02 



Project Description: 

Objectives; ,._ ..^ , ., ■ . _ , ,. • ? . -, 

Current methods of interactive external beam treatment planning are 
purely trial and error. A more rational approach should be developed, 
with the computer aiding in optimization. 

Methods Employed: , - _, , ,,. .''...- , ,, ',,, j ; 

The programmed console is used to plot isodose patterns. Linear 
programming optimization techniques are used for fast response. 

Major Findings: 

A more directly interactive system utilizing optimization has been 
designed for external beam treatment planning. The operator is enabled 
to set dosage constraints directly into a display of the patient's 
cross-section. Linear programming optimization chooses beam strengths for 
uniformity within the constraints. The operator is thereby enabled to try 
many more beam directions and wedges than possible under current methods. 
The system has been simulated on examples of treatment plans. In one 
case a significant improvement over a previously used plan was discovered. 



154 



Significance to Biomedicine: 

Radiation therapy is widely used for cancer patients. About one out of 
five requires individual treatment planning. The proposed system would 
be able to aid in this task. 

Proposed Course: 

Various avenues for implementing the system will be investigated and an 
appropriate route taken. 



155 



Serial Number DCRT 6.80 

1. Heuristics Laboratory 

2. 

3. Bethesda 

PHS - NIH 
Individual Project Report 
July 1, 19 71 through June 30, 19 72 

Project Title: Information Storage and Retrieval with Index 

Previous Serial Number: None 

Principal Investigator: Lewis M. Norton, Ph.D. 

Other Investigators: Robert Magnuson, DMB, DCRT. 

Cooperating Units: DMB, DCRT. 

Man Years: 

Total: 0.15 

Professional: 0.15 
Other: 0.0 

Project Description: 

Objectives: 

To extend the usefulness of DCRT software systems for infoinnation storage 
and retrieval, by providing a retrieval capability based on an index of 
textual material in files. 

Methods Employed: 

The RMAG language is used to create user-specific software for indexing a 
file and querying the index using Boolean combinations of keywords. 
Sophisticated OS data management techniques are employed. The routines 
are interfaced with existing DMB software for file creation and maintenance. 

Major Findings: 

None 

Significance to Biomedicine: 

This system provides a fast, flexible means of searching files containing 
data in textual form which can be used in a variety of applications. 

Proposed Course: 

Modifications and/or additions will be made to the system to meet the needs 
of new applications. 

156 



A simple example was discovered of a non-terminating program which can fool 
an unsophisticated program verifier. 

A method was developed for improving turn-around time by allowing users some 
control of the priority of their run. 

Significance to Biomedical Research: 

As biomedicine uses more and more sophisticated programming techniques , 
many of these findings will be important. Specifically, automatic pro- 
gramming aids will allow a scientist to use computers effectively without 
becoming a computer specialist. 

Proposed Course : 

Work on refinement of partial evaluation and automatic program synthesis 
are established projects which may be set aside as other tasks become 
more urgent. Other applications to programming theory and practice are 
serendipitous by-products of our work. 

Publications : 

Jones, T. L. : Load management for short turnaround. Proceedings of the 
SIGCOSIM Second Annual Technical Symposium (to be published). 



157 



Serial No. 



DCRT 6.81 



1. Heuristics Laboratory 

2. 

3. Bethesda 

PHS - NIH 
Individual Project Report 
July 1, 19 71 through June 30, 1972 

Project Title: Microbiology Data Bank 

Previous Serial Number: None 

Principal Investigator: Lewis M. Norton, Ph.D.; Micah J. Krichevsky, Ph.D., NIDR. 

Other Investigators: None 

Cooperating Units: Environmental Mechanisms Section, NIDR. 

Man Years • : 



Total: 


0, 


,2 


Professional: 


0, 


,2 


Other: 


0, 


,0 



Project Description: 

Objectives: 

To develop software enabling on-line and batch coding and retrieval of 
data about bacteria and other microorganisms. 

Methods Employed: 

Systems analysis and software development is being done to meet the unique 
requirements of the international community of microbiologists, in order 
that the maximum amount of data about microorganisms may be accumulated and 
made available. The kinds of data as well as their uses are extremely diverse. 

Major Findings: 

We are encoding data on antibiotic sensitivities of over 9000 strains of 
pathogenic bacteria from NIH's Clinical Pathology Laboratory, CC. NIDR is 
encoding data on oral microorganisms. Groups from Australia, England, Japan, 
Czechoslovakia and many other countries, as well as from many points in the 
United States, have indicated interest in contributing to and making use 
of such a data bank. 



158 



Significance to Biomedicine: 

Data about bacteria and other microorganisms is of importance to many 
medical specialities. 

Proposed Course: 

The coding method will continue to be expanded to allow for additional 
types of data. On-line aids to using the system will be designed and 
implemented. Use of the data bank will have to be possible from 
different types of computers. Increasingly sophisticated retrieval methods 
for utilizing the data will be developed. 



159 



Serial No. DCRT 6.90 



1. Heuristics Laboratory 

2. 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Nuclear Magnetic Resonance 



Part A 
Part B 
Part C 



NMR Search Program 
Graphical NMR Analysis Program 
C NMR Calculations 



Previous Serial Number: None ~ -^--: a;.-^ . 
Principal Investigator: Stephen R. Heller, Ph.D. 



Other Investigators: 



Cooperating Units: 



Man Years 



Richard J. Feldmann ,CCB; Arthur E. Jacobson, Ph.D., 
LC:NIAMD; Robert J. Highet , Ph.D., LC:NHLI; Edward A. 
Sokolowski, LC:NHLI 

CCB; Laboratory of Chemistry, NIAMD; Laboratory of 
Chemistry, NHLI 



Total: 

Professional: 

Other: 



Part A: NMR Search Program 

Project Description: 

Objectives: 

The development of an instantaneous Nuclear Magnetic Resonance (NMR) 
spectral search program. 

Methods Employed: 

The chemical shifts and structural representation for the 742 compound 
Varian NMR catalog file was undertaken. A computer program for the PDP-10 
was written to search for chemical shifts from the file. 

Major Findings: 

Searching of the file can be performed in an instantaneous interactive 
fashion on the PDP-10. 



160 



Significance to Biomedicine : 

The ability to have a library of NMR spectra readily available from com- 
puter terminals at the NIH will aid scientists using NMR as a tool for 
structural elucidation of biologically significant molecules. 

Proposed Course: 

WLN, chemical names, molecular formula and molecular weights will be 
added to the file and the programs rewritten to incorporate this added 
information. 

Publications: 

Heller, S. R. and Feldmann , R. J.: An Interactive NMR Chemical Shift Search 
Program, J. Chem. Ed . 49: 291, 1972 

Part B: Graphical NMR Analysis Program 

Project Description: 

Objectives: 

To interactively analyze nuclear magnetic resonance (NMR) spectra for the 
chemical shifts and coupling constants contained in the spectra. 

Methods Employed: 

An interactive NMR analysis program has been written for the PDP-10 which 
allows for the graphical display of calculated NMR spectra. The program 
also contains provisions for two types of hard copy Calcomp output which 
are useful for permanent records and publications. 

Major Findings: 

Analysis of NMR spectra of complex spin systems are readily handled in a 
graphical interactive mode. The overall elapsed time for analysis has 
been reduced from weeks to days using computer graphics. 

Significance to Biomedical Research: 

This program is believed to be the most sophisticated and comprehensive 
NMR analysis program available. The combination of computer graphics and 
hard copy plotting options is unique. 

Proposed Course: 

As the program is used more, it is anticipated that various users will re- 
quest additional options be added, such as simultaneous display of experi- 
mental and calculated spectra. 



161 



13 
Part C: C NMR Calculations 

Project Description: 

Objectives : 

13 
The automatic computer calculation of C Nuclear Magnetic Resonance 

Chemical Shifts. 

Methods Employed: 

13 
Using tables of pioblished C NMR chemical shifts, a correlation of 

structure and shifts are used to obtain a set of coefficients for the 

computer calculation of other C chemical shifts. 

Major Findings: 

13 
A program has been written which can accurately calculate the C chemical 

shift for any straight-chain hydrocarbon. 

Significance to Biomedical Research: 

13 

The ability to calculate spectral properties such as the C chemical 

shifts, are very valuable in structure determination. 

Proposed Course: 

To expand the program to calculate shifts for cyclic compounds, compounds 
with aromatic systems and compounds with hetero-atoms . 



162 



Serial No. DCRT 6.91 

1. Heuristics Laboratory 

2. 

3. Bethesda 

PHS - NTH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Mass Spectrometry Data Retrieval System 

Previous Serial Number: None 

Principal Investigator: Stephen R. Heller, Ph.D. 

Other Investigators: Henry M. Fales , Ph.D., G. W. A. Milne, Ph.D., 

David C. Redder, Ph.D., LC:NHLI 

Cooperating Units: Laboratory of Chemistry, NHLI 

Man Years 

Total: .9 

Professional: .8 
Other: .1 

Project Description: 

Objectives : 

The development of an interactive conversational search system for mass 
spectrometry data. 

Methods Employed: 

A highly conversational, Interactive search system for a large file of 
mass spectral data was written for the PDP-10. The system has eight op- 
tions . They are : 



Peak and Intensity Search 

Molecular Weight Search 

Complete Molecular Formula Search 

Partial Molecular Formula Search 

Molecular Weight and Molecular Formula Search 

Molecular Weight and Peak Intensity Search 

Molecular Formula and Peak/Intensity Search 

Spectrum Printout 



Major Findings : 

It is possible to Instantaneously search an 8124 mass spectral file with 

163 



resulting answers identifying the molecule or suggesting possible solutions. 
The search system can be readily accessed from any computer terminal at 
NIH, and is especially useful in the laboratory with a mass spectrometer. 

Significance to Biomedical Research: 

The ability to have a large library of mass spectra always available with 
a computer terminal and time-sharing PDP-10 computer has enabled scientists 
to readily identify and assist in the structural elucidation of biologically 
significant molecules. The system has greatly aided the scientists in- 
volved in the mass spectral analysis of blood samples from drug overdose 
patients from local hospitals. 

Proposed Course: 

It is planned to expand the size of the file with the emphasis on adding 

drugs, pesticides and related chemicals of biological interest. Also, 

structural information will be added to the file so that the system search 
options can be expanded. ; 



164 



Serial No. DCRT 6.92 



1. Heuristics Laboratory 

2. 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Energy State Calculations 

Previous Serial Number : None 

Principal Investigator: Stephen R. Heller, Ph.D. 



Other Investigators : 



Cooperating Units: 



Man Years 



Jack S. Cohen, Ph.D., PSL; Richard J. Feldmann, CCB; 
Arthur E. Jacobson, Ph.D., LC:NIAMD; Ira A. Levin, 
LPB:NIAMD 

Physical Science Laboratory; Computer Center Branch; 
Laboratory of Chemistry, NIAMD; Laboratory of Physical 
Chemistry, NIAMD 



Total: 

Professional: 

Other: 



Project Description: 

Objectives: 

The development and use of computer programs for calculating the conforma- 
tion and energy of a molecule. 

Methods Employed: 

A number of computer programs to do energy minimization and energy state 
calculations has been investigated. The "CNDO/INDO energy state calculation 
program has been redesigned and rewritten for the PDP-10 and is to be used 
in conjunction with X-Ray Modeling System on the PDP-10. 

Major Findings: 

There are no computer programs available for energy minimization of systems 
of biological interest. If a conformation is known, the CNDO/INDO program 
is a reasonable program to use for the calculation of that conformation. 



165 



Proposed Course: 

To undertake a further study of the methods used to minimize energy states 
in an attempt to write a computer program for use with biologically in- 
teresting molecules. 



166 



Serial No. DCRT 6.93 

1. Heuristics Laboratory 

2. 

3. Bethesda 

PHS - NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Retrospective CBAC Literature Search 

Previous Serial Number: 2.16 

Principal Investigators: Stephen R. Heller, Ph.D, Richard J. Feldmann, CCB 

Other Investigators: None 

Cooperating Units: CCB 

Man Years : 

Total: .2 
Professional: .2 
Other: 

Project Description: 

Objectives : 

The development of an Instantaneous retrospective search capability of 
the CBAC (Chemical Biological Activities Abstracts) published by Chemical 
Abstracts Service. 

Methods Employed: 

Scatter storage techniques (hashing) were investigated and applied programs 
were written and debugged for : 

1. Processing CBAC tapes 

2. Generating a HASH file and 

3. Retrieving CBAC references from the HASH file 

Major Findings: 

An Instantaneous retrospective search of CBAC for the occurrences of a 
term (compounds, REGN (registry numbers), authors, etc.) can be performed. 
The Interactive analysis of search results by the user leads to a more 
complete utilization of the data base and the retrieval results. The 
present Implementation obtains adequate coverage of this abstract journal 
by storing only the reference (volume, issue, abstract number) to the CBAC 
abstract. 



167 



significance to Biomedical Research: 

When all the sections of chemical abstracts are produced by computer the 
CBAC format will be used. The experience gained with CBAC will lead to 
a more complete computer coverage of chemical literature. 

Publications: 

Heller, S. R. , Feldmann, R. J. and Shapiro, K. P.: An Experimental Com- 
puterized CBAC Search Project, J. Chem. Doc. 11: 248-251, 1971. 



168 



July 1, 1971 through June 30, 1972 

PHS-NIH 

Division of Computer Research and Technolo/.y 

Summary of Branch Activities 1. DCRT-7 

7. Data Management Branch 3. J. Emmett word 

Branch Chief 

I . SUMMARY 

The Data Management Branch Is responsible for providing the NIH 
with efficient/ effective and easily accessible computer based 
capabilities for data management and analysis. 

During fiscal year 1972, many new applied systems were developed 
and implemented to assist the NIH research and administrative 
activities. Significant among these systems were the file 
management system developed to support the Type 1! Coronary 
I ritervent I on Study for the National Heart and Lung institute 
and final modification and implementation of phase one of the 
Care inogen I si s Bioassay Data System in support of the National 
Cancer Institute. The conversion of the Clinical Center's 
Laboratory Beta file was also completed during this period. 
This conversion now enables us to provide the research In- 
vestigator with an easily accessible data base for analysis. 
To date several patient subfiles of laboratory data have been 
created for individual analysis and it is anticipated that 
the demand for this data will continue to grov;. 

The Cancer survival system was generalized during the year 
to provide other researchers with its capability to display 
survival statistics of subjects Involved in a research 
protocol. Extensions to this system Included such thing as 
median survival time calculations and the option to request 
a point plot of data at various levels within the system. 
The result of these extensions has been the expansion of the 
former esoteric use of the system by the Cancer End Results 
Section to Include both other groups in the National Cancer 
Institute and other institutes at the NIH such as the 
National Heart and Lung Institute. 



169 



The Mathematical and Statistical Program Manual was completely 
revised and updated during the year. In order to facilitate 
access to the library programs, the entire library Is nov; 
being moved onto a private disk pack. This approach enables 
the branch to provide programs to the user whicli are ready to 
be used immediately and also eliminates other problems related 
to storing library programs and subroutines on public libraries, 
This new library is currently being tested in our daily produc- 
tion environment and it is anticipated that this v, i 1 1 become 
our primary mathematical and statistical library by early 
fiscal 1973. in addition to the regular programs the Branch 
is now offering the I iiM Continous System liodellng Program to 
i ts users . 

The Clinical Center's interview scheduling system required a 
major revision to enable scheduling of intervlevvs for potential 
clinical associates over a two day period. Cesplte manpower 
limitations the branch vyas able to provide a completely re- 
vamped system which was used to schedule interviews in 
fiscal year 1972. 



in i 
the 
prov 
s ys t 
ret r 
proc 
I nc 1 
add i 
the 
I n te 
prog 
beg I 



ts con 1 1 n 
Data Man a 
ide sourc 
cms to cr 
ieve data 
ess i ng . 
uded in t 
tion to t 
need for 
ract I ve t 
ram libra 
n Impleme 



ued us 
gemen t 
e lang 
eate a 

f r om 
The fi 
h e D a t 
his, w 
table 
ermi na 
ry . I 
n ta t i o 



e of 
Bra 
uage 
nd u 
f I le 
na 1 
a f ia 
e ar 
gene 
1 ac 
f an 
n of 



the 
nch 

gen 
pdat 
s ei 
docu 
nage 
o no 
rato 
cess 
a 1 ys 

the 



Rec 
has 
era t 
e fi 
ther 
ment 
men t 
w i n 
rs, 

to 
i s p 
se c 



u rs I V 
compl 
i on t 
les, 

by w 
at ion 
Bran 
the 
and t 
them 
roves 
apab i 



e Mac 
eted 

ass 
to re 
ay of 

of t 
ch Ge 
proce 
he po 
athen 

f avo 

1 i tie 



ro Act 
proj ec 
i s t In 
format 
termi 
hese s 
nera to 
ss of 
ss i b 1 i 
at i ca 1 
reb 1 e, 
s . 



uated Gene 

ts which 

programmi 

files, an 

na 1 or bat 

ys tems wi 1 

rs f "anua 1 . 

analyzing 

ty of prov 

and s ta t i 

the branc 



rator, 

ng of 
d to 
ch 

1 be 
In 
both 
I d i n g 
St i cal 
h w I 1 1 



170 



Serial No. 7.1 

1. Data Management Pranch 

2. Appl led Systems 

Programming Section 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1^ 1971 through June 30, 1P72 

Project Title: Cigarette Condensate Study 

(Phase 2) 

Previous Serial Number: 7.1 

Project Leaders: Mrs. C. Staneck, PCRT 

Mr. n. Dobenecker, nCRT 

Other Professionals: None 

Cooperating Units: Etiology Pranch, NCI 

Biometry Branch, NCI 

Han Years: 

Total: O.Uit 

Profess ional : f). 39 
nther: n.05 

Project Description: 

Object i ves : 

The purpose of the system, as in Phase 1, was to 
collect and report on information gathered by the 
Hazelton Laboratories. The information pertained 
to the occurences of tumors (size, number, kind, 
etc.) in 6,500 animals treated with applications 
of a nicotine type solution in varying concentra- 
tions. The specifications of the solution were 
determined by Dr. Oori of the National Cancer- 
institute. In addition a statistical analysis, 
under the direction of Pr. John Hart, NCI, has been 
undertaken to evaluate the condensates used in 
Phase 1. 



171 



Methods Employed: 

The programs written for the collection of data in 
Phase 1, (i.e., edit, update, reports) which pro- 
cessed 7,100 animals over a period of eighteen 
months were, with minor modifications, used for the 
data collection In Phase 2. 

Programming for the Phase 1 data analysis has in- 
cluded a Chi Square analysis of cage effects, sur- 
vivorship tables, and plots for comparing survival 
among condensates. Programmers in the Biometry 
Branch, NCI have prepared rank order tests and curve 
fitting of the survival rates. 

flajor Findings: 

Analysis of the Information resulting from these 
computer programs is being done by the Mathematical 
Statistics and Applied Mathematics Section, Biometry 
Branch, NC I . 

Sign I f ance : 

A review of the data supplied by the analyses is 
in progress. 

Proposed Course: 

Continued programming support for this project will 
be provided as it Is required. The programs pre- 
pared for the analysis of Phase 1 are expected to 
be used in the analysis of the data collected In 
Phase 2. 

Honors and Awards: None 

Publ Icat ions : Mone 



172 



Serial No. 7.? 

1. Hata Management branch 

2 . Appi led Systems 

Programming Section 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, iq71 through June 30, 1P7? 

Project Title: Carcinogenesis Pioassay Hata System 

Impl ementat ion 

Previous Serial Number: 7.2 

Project Leaders: Mary F. Linhart, nCPT 

Robert L. Martin, nCRT 

Other Investigators: Anne M. Gallagher, ncRT 

John R. Parks, nCRT 
Sandra Foote, PCRT 
Roger H. Pailey, nCPT 
Heorge K. Dobenecker, nCRT 
Charles I. Twigg, nCPT 

Cooperating Units: Program and Data Analysis I'nit, 

Office of Associate Pirector for 
Carcinogenesis, Ftiology Area, 
NCI 

Man Years: 

Total: 2.10 
Professional : 2.05 
Other: 0.05 

Project rescript ion: 

Object i ves : 

Implement the Carcinogenesis Rioassay Pata System 
which will facilitate contract administration and 
permit more satisfactory evaluation of experiments. 
Provide computerized editing and file maintenance 
procedures. Pevelop a basic format to structure 
reporting of test results in all investigations. 
Pevelop procedures for user handling of input and 
systems operation. Investigate and analyze addi- 
tional reporting procedures. 



173 



flethods Employed: 

Standard Information systems design and programming 
techniques have been used. 

flajor F i nd ings : 

The Carcinogenesis Bioassay Data System has alerted 
the Program and Pata Analysis Unit (PDAU) of NCI 
that one of the bioassay contractors has not been 
performing complete pathology on all the animals as 
is required by the contract. This error by the 
contractor vvould have rendered his experiments use- 
less. The PDAU claims that this incident alone has 
made the system pay for itself. Further, the system 
has shown that certain contractors have not set up 
their experimental designs properly. In these 
cases, the reports issued by the system alerted the 
PDAU early in the contract period enabling them to 
contact the contractor so that he could correct or 
eliminate the problem. 

Significance to Bio-medical Research and the Program of 
the I nst I tute : 

The Carcinogenesis Bioassay Data System is intended 
to facilitate analysis of experimental results on 
large numbers of animals and tests by imposing 
structure upon massive collections of data, so that 
comparisons and correlations can be made upon 
experimental results within an investigation and 
among multiple Investigations. It is expected that 
establishment of files will allow for more sophi- 
sticated analysis of experimental results. 

Proposed Course: 

File maintenance and a basic reporting system have 
been programmed, tested and are in full operation. 
A checkpo i n t- res ta rt system has been designed and 
is under test. A file compatibility system is 
being developed to guarantee file integrity. 

national Cancer Institute will continue to be re- 
sponsible for data collection, update and reporting 
schedul i ng. 



174 



nCRT v;Ml continue to provide maintenance anH 
support of the basic system. In addition other 
programs will be developed as desired by the 
Program and Data Analysis Unit of the National 
Cancer Institute to correlate data and report 
f ind ings . 



Honors and Awards: None 



Publ I cat Ions : 



None 



175 



Serial No. 7.5 

1. Data Management Branch 

2 . Appl ied Systems 

Programming Section 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: NIAID Research Contracts 

Previous Serial IJumber: 7.3 

Project Leader: Judy Mahaffey, DCRT 

Other Professionals: David Blessley, DCRT 

Cooperating Units: NIAID 

Man Years : 

Total: 0.2 

Profess ional : 0.15 
Others: 0.05 

Project Description: 

Object i ves : 

The purpose of this project was to design and imple- 
ment a system to 1) organize research contract data 
into a computer accessible file 2) provide file up- 
date capabilities 3) prepare the necessary recurring 
management reports and k) provide the capability of 
retrieving, in an organized fashion, data needed to 
meet one-time management requirements. 

Methods Employed: 

A WYLBUR system with IRS capabilities has been de- 
signed and implemented. The system makes use of both 
batch and on-line processing. All research contracts 
data is entered directly into the system through a 
2741 terminal and stored on a V.'YLBUR data set. A 
COBOL program is then run periodically to edit and 
re-format this data for later processing. Errors 
identified by the COBOL program are corrected on- 
line. After the update cycle is complete, IRS and 
COBOL programs are submitted to prepare a variety of 
management reports. 



176 



Major Findings and Significance: 

The system now serves as a powerful data management 
tool for NIAID. It stands as proof that a single 
non-data processing person can maintain and operate 
a computer system if the system is so tailored to the 
user. Perhaps the primary significance of the system 
Is that all research contracts data can he maintained 
and selectively invoked for any one or all ?h reports 
options by submitting a single computer run. In 
twenty minutes to two hours the requested reports 
are available to management. All time previously 
spent on manual data verification and report pre- 
paration is freed for analysis tasks. 

f^roposed Course: 

The complete system and documentation package have 
been turned over to the Contract Management ^'ranch, 
M I A I D . 

Honors and Awards: None 

Publ i cat ions : None 



177 



!^erfa1 Mp . 7 . tt 

1. Data Management ^ranch 

2 . Appl led Systems 

Programming Section 

3. Rethesda 

PHS-MIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Collection of Pleeding Data Related to 

the Study of Kuru 

Previous Serial Number: 7.k 

Project Leader: George Hobenecker, nCRT 

Other Professionals: Dr. Steven '^Hesenfeld (NINDS) 

Cooperating Units: NINDS Collaborative and Field Research 

Branch 

Man Years : 

Total: n.2n 
Profess ional : 0.10 
Other: 0.10 

Project r>escr ipt ion : 

Object i ves : 

The purpose of the project was to take bleeding data 
collected by Pr. C. Oajdusek and build a data base 
that would be appropriate for extensive computer 
analysis of the data. 



178 



Methods Fmployed: 

Forms were designed so that the information could be 
coded and then punched into a machine readable format, 
An edit and update program was v/rltten. The system 
was supported by the standard update facility v/hereby 
records can be inserted, modified and/or deleted just 
by entering the proper update code (l^t*,'^) and the 
key of the record in question along v/i th the desired 
data. Report programs were v^ritten to produce 
listings of the file on request. 

Major F i nd I ngs : 

None . 
!^Ign I f icance : 

None . 

Proposed Course: 

Support any additional request by Hr. f^ajdusek or 
Hr. '-'iesenfeld In the analysis of the data. 

I^^onors and Awards: None 

Publ icat ions : None 



179 



Serial No. 7.5 

1. Data Management Branch 

2. Office of the Chief, 

DMB 

3. Bethesda 

PHS-[iill 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Consulting and Tutorial Services 

Previous Serial ilumber: None 

Project Leader: J. M. S. Prewitt, PCRT 

Other Professionals: 

Cooperating Units: National Institutes of Health:- 

CCB/DCRT, OD/nCRT, LAS/DCRT, 
;JCI/CHEM, :JinMS/RG, NIMH/AP, 
illCH/EB. 

National Science Foundation:- Office 
of Computing Activities 
I.E.E.L. and Engineering Foundation 

"an Years : 

Total : 0.15 ; ' . 

Profess ional : 0.15 
Other: 0.00 

Project Description: 

Objective: 

To disseminate and exchange information on applied 
mathematics and computer applications (including 
modeling and picture processing) as they relate to 
data analysis in biomedicine, and to facilitate such 
exchanges . 

f.ethods Employed: 

Technical consultation and evaluation is provided in 
response to direct requests from biomedical researcher; 
at :.'.I.!^. and other institutions. Courses, seminars 
and lectures are arranged to meet local N.I.H. needs. 



180 



Major FI nd i ngs : 

Technical consultation and teaching Is needed for 
diverse data analysis and management problems. 

Significance to Riomedical Research: 

Information exchange and scientific liaison are 
obviously fundamental to progressive research and 
appl i cat ions . 

Honors and Awards: None 

Publ Icat Ions : Mone 



181 



Serial No. 7,6 

1. Data Management Branch 

2. Office of the Chief 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Bioassay Data Collection System 

Previous Serial Numbers: None 

Project Leaders: William D. Vincent 

Joseph Campbell 

Other Professional: None 

Cooperating Units: Carcinogenesis 

flan Years: 

Total: 0.60 
Professional: 0.60 
Others: 0.00 

Project Description: 

Obj ect i ves : 

1. To develop a data collection system that collects 
and stores, in transaction formats, experimental 
data submitted to NCI from contractors in the 
Bioassay Program. 

2. To determine the feasibility of digitizing data 
using a real-time system. 

3. To study the applicability of using data collec- 
tion system to train field personnel in the pre- 
paration of experimental data at the contractor 

1 eve 1 . 

flethods Employed: 

A system of programs was written In CPS that prompts 
the operator for specific parameters, edits some of 
the parameters, and stores them in transaction 
formats for further procession by the CBDS system. 
The operator of the data collection system has the 
capability of modifying transactions before they are 
written to an on-line file. 

182 



Major F I nd i ngs : 

The notion of developing an interactive data collec- 
tion system is certainly a practical and productive 
way of digitizing raw data. Specifically, a real- 
time systems is an excellent way of minimizinp the 
interface between man and computer. 

S i gn i f i cance : 

None 

Proposed Course: 

NCI will be responsible for operating the data 
collection system. DMB will provide personnel for 
maintenance and support of the system. 

Honors and Awards: None 



Pub 1 i cat ion ; 



None 



183 



Serial lio. 7.7 

1. Data Management Branch 

2. Office of the Chief, 

DMB 

3. Bethesda 

PHS-;iiH 
Individual Project Report 
July 1, 1971 through June 50, 1972 

Project Title: Characterization of Normal and 

Abnormal Bladder Epithelium by 
Digi tal Computer 

Previous Serial Number: None 

Project Leader: J. S. Prewi tt, DCRT 

Other Professionals: G. Friedell, M.D., Chief of Pathology 

St. Vincent's Hospital, Worcester, flass 

R. Malmgren, M.D., NCI 

E. Soto, M.D., St. Vincent's f^ospital, 

V/orcester, Mass. and Boston University 

Medical School 

n. K. Reece, HL/DCRT 

Man Years: 

Total: n.20 
Professional : 0.20 
Others: 0.00 

Project Description: 

Object i ves : 

The objectives of this project are (1) to obtain 
quantitative characterization of normal and ab- 
normal bladder epithelium from digitized micro- 
scans of stained sections, (2) application of 
characterizations obtained for cells in s i tu to 
exfoliated cells, (3) evaluation of the diagnostic 
and prognostic potential of these characterizations, 
and (it) development of computer support for data 
acquisition, analysis, and representation. 



184 



Methods Fmployed: 

Characterization involves the development of new 
image processing techniques for handling complex 
textural pictorial material such as tissues, as 
well as statistically-based diagnostic and 
prognostic rules. 

Major Findings: 

The project is in the formulation stage, and is 
awaiting funding. 

S ign i f icance : 

Ouantitative characterization of exfoliative and 
histologic material from the urinary tract may lead 
to new methods for early detection of cancer which 
may be performed routinely and with no discomfort to 
the patient. In addition, digital mi croscann i ng 
coupled with computer analysis, may provide a 
unique, reliable method for studying relative activ- 
ity rates and biochemical specificity in normal and 
cancer eel Is. 

Proposed Course: 

A formal proposal for this study is being prepared 
in response to an external request from the National 
Advisory Committee on Cancer of the Pladder, and 
calls for a mul t i - Inst i tut ional , mul t i -di sclpl inary 
collaborative effort. If approved, a feasibility 
study to determine characterizations for bladder 
epithelium based on both histologic and cellular 
material will be pursued in the fiscal year l'^72-7?. 

Honors and Awards: None 

Pub! icat ions : 

Prewitt, J.S.: Biomedical Picture Processing: 
Synopsis and Challenge. Invited Paper, Fng I neer Ing 
Foundation Research Conference on Pattern Information 
Processing Systems , Airlle f^ouse, Warrenton, Virginia, 
1972. 



185 



Serial No. 7.8 

1. Data Management Branch 

2. Software Support 

Sect i on 

3. Bethesda 

PHS-NI H 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Generated Update Facility 

Previous Serial Number: 7.8 

Project Leader: Darius Georg, DCRT 

Other Professionals: Robert A. Magnuson, DCRT 

Cooperating Units: None 

Man Years: 

Total: 2.00 

Professional : 2.00 
Others: 0.00 

Project Description: 

Obj ect I ves : 

Provide an update and edit program such that the 
generated source language program Is a tool for 
establishing and maintaining a computer file with 
virtually no programming effort. Minimize the 
time and effort of creating and maintaining a 
master f I le of records. 

Methods Employed: 

An RMAG language program is used to generate an 
update and edit program In one module. The update 
program uses a naming scheme for updating fields. 
The output is a USA COBOL PROGRAM, documentation 
of the master record, an error list of invalid 
arguments used on the Input parameter cards, and 
JCL needed to execute the program after it has 
been comp I 1 ed . 



The following is a list of features of this module: 

1. Update by file name 

2. Free form input (space delimeters) 

3. Built in transaction formatting 
k. Update messages and error report 

5. Accounting information for each run 

6. A listing of the input transactions 

7. Update by groups. 

Proposed Course: 

The system be extended for a larger set of re- 
quirements providing a wider range of action to be 
performed during update execution. 

The following is a list of suggestions to be con- 
sidered for future implementation of the update 
and od i t system. 

1. Provide for field accumulation if desired 
rather than writing over an existing value in 
a record. 

2. Have the ability to process MASS changes on u 
1!^! 1 „- -.11 basis. 



3. 
k. 
5. 



conditional or all basis. 

The ability to process variable length 

Provide a decimal capability. 

Consider a means of alleviating the CORE limi 

tation problem associated with large records 

and if feasible, provide a means for handling 

repeti ti ve codi ng. 



records, 
CORE 1 imi - 



Honors and Awards: None 



Publ i cat ions : None 



187 



Serial No. 7.9 

1. Data Management Branch 

2. Software Support 

Sect I on 

3. Bethesda 



PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Recursive Macro Acutated 

Generator 

Previous Serial Number: 7.9 ^ ^ 

Project Leader: Robert A. Magnuson, DCRT 

Other Professionals: None 



Man Years: 




Total : 


0.75 


Profess ional : 


0.75 


Others: 


0.00 



Project Description; 
Obj ect i ves : 



Provide a powerful programming tool capable of 
generating source programs in any and all pro- 
gramming languages. 



Methods Employed: 



Continuing the project from the preceding year, 
the stand-alone macrof;enera tor was improved. Im- 
portant new features v;ere added and the execution 
time was reduced by one-half. The library of RMAG 
subprograms vjas revised to take advantage of the 
new language features, and new subprograms vjere 
added. The documentation was revised and reissued, 



188 



Major Findings 



A number of RMAG programs (both newly written and 
modified prior ones) became standard Data Manage- 
ment Branch products, each capable of generating 
a customized source language program. The number 
of analysts using these DMB generator packages grew 
significantly. Also the number of people writing 
their own RMAG programs slowly grew. Universities 
and other government computer installations re- 
quested and are using the RMAG system. 



S i gn i f i cance : 




Proposed Course: 

The system will be extended and more broadly 
appl i ed. 

Honors and Awards: 

While not a formal honor, we were pleased that the RMAG 
system is being used at Princeton University Computer 
Center. They have included our RMAG manual as part of 
their Computer Center Guide. 

Publ i cat ions : 

Technical Report No. 7, Recursive Macro Actuated 
Generator, DCRT, July 1971 and revised March 1972. 



189 



Serial No. 7.10 

1. Data Management Branch 

2. Scientific Applications 

Sect i on 

3. Bethesda 



PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 



Proj ect Title: 

Previous Serial Number 
Project Leader ( s ) : 
Other Professionals: 



Cooperating Units: 






Man Years: 






Total: 


0. 


,15 


Profess ional : 


0. 


,15 


Others: 


0. 


,00 



Current Awareness Search of 
Chemical Literature 

7.10 

Mary Lee Dante, DCRT 

Myrtl e Morr I s, DCRT 
Georg Schulze, DCRT 

None 



Project Description 
Obj ect i ves : 



Provision of a bi-weekly current awareness search- 
ing of the CBAC ( Chemi ca 1 -B I ol og i ca 1 Activity) 
journal published by Chemical Abstracts Service. 



Methods Employed 



Tapes received bi-weekly from CAS are converted to 
MARC format using DCRT-mod I f i ed programs from the 
National Science Library of Canada. The tape is 
then searched and resulting abstracts (those which 
match profiles or search requests) are mailed to 
CBAC subscribers. 



Major Findings 



It has been found economically non-feasible to 
charge for this service. 



190 



Significance: 



Provides chemists at NIH with a searchable data 
base of selected literature on chemical compounds 
of biological interest. 

Proposed Course: 

CBAC will be offered as a free service to the NIH 
Community. A retrospective service covering the 
period from 1965-1971 will be set up on a monthly 
basis. _ The possibility of additional data bases 
c^'"^ in?!'"'^?'^ '" ^^^ service is being explored. 
Since CBAC will be available only in the Standard 
Distribution Format beginning in January, 1973, 
It will be necessary to provide a new reformat' 
program. If it is possible to obtain a PL/1 com- 
piler which produces more efficient code than the 
present compiler, it will be necessary to modify 
and update the new reformat programs provided by 
the National Science Library of Canada. Otherwise 
a new program will be written in assembly language! 

Honors and Awards: None 

Publ icat ions : None 



191 



Serial No. 7.11 

1. Data flanagement Branch 

2. Office of the Chief, 

DMB 

3. Bethesda 



PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 



Project Title: 
Previous Serial Number: 
Project Leader: 
Other Professionals: 

Cooperating Units: 

Man Years : 

Total: n.20 
Profess ional : 0.20 
Others: 0.00 

Project Description: 

Objectives: 



Genetics of Lipoprotein Metabolism 

None 

J. S. Prewitt, DCRT 

D. Fredrickson, M.D. NHL I 
R. Levy, M.D., NHL I 



The objective of this project was to advise Dr. 
Fredrlckson on the applicability of advanced mathe- 
matical and computer techniques to research pro- 
grams within his purview. The principal problem 
area concerns genetically determined control systems 
for lipoprotein, cholesterol, and triglyceride 
me tabol i sm. 



192 



Methods Employed: 

Exploratory discussions were held to get a feeling 
for research needs and the extent to which they are 
now being met by DMB. 

Major Findings: 

implicit decomposition of univariate patient data 
into one, two or three classes, representing homo- 
zygous non-affected, heterozygous affected, and 
homozygous affected subjects, respectively, could 
be handled by IJHL ! using elementary techniques and 
desk calculation. A formal decomposition Into sub- 
samples Including estimation of population means 
and covarlances and relative proportions, with confi 
dence Intervals, was suggested to NHLI, but they 
felt this to be more than they needed for Immediate 
purposes. However, this technique is of Interest 
from the viewpoint of applied mathematical sta- 
tistics, and a nonsuperv i sed learning algorithm was 
pi anned . 

Signl f I cance : 

NHLI is engaged In extensive data collection on 
dysllpldemla and Its control, and regards their 
effort as a major committment. 

Proposed Course: 

Consultation services will be provided to NHLI as 
requested. 

Honors and Awards: None 

Publ icat Ions : 

Prewi tt, J .S. : 

Biomedical Applications of Nonsuperv I sed 

Pattern Recognition. 

Invited paper. Fourth IEEE Pattern Recognition V/orkshop, 

liot Springs, Virginia, 1972. 



193 



Project Title 



Serial [Jo. 7.1? 

1. Data Management Pranch 

2. Office of the Chief, 

DMB 

3. Bethesda 

PHS-;jin 

individual Project Report 
July 1, 1971 through June 30, 1972 

; A Decision-Theoretic Approach to 

Optimizing Radiotherapy 



Previous Serial IJumber: 
Project Leader: 
Other Professionals: 

Hoope ra t i ng Un i ts : 

fan Years : 

Total : 0. 20 
Profess ional : 0.20 
Others: 0.00 

Project Description: 



None 

J. S. Prewitt, DCRT 

L. A. Davis, fl.D., Department of 
Radiology, University of 
Pennsylvania Hospital 

None 



Obj ect i ves 

The o 
the c 
rad i a 
mot i V 
gramm 
mi t e 
(i i ) 
ter I a 
ent i a 
and n 
and r 
V ide 
progr 
(via 
rad io 



bjectives of this project are: (1) to replace 
urrent primitive empirical approaches for 
t i on treatment planning with a rad i ob i ol og I cal 1 y 
ated approach, (2) to provide a flexible pro- 
ing system for such an approach which v/ill per- 
xpansion and alteration of (i) a priori data, 
dosimetry calculations, (ill) optimality crl- 
, and (iv) mathematical models for the differ- 
1 grov^th kinetics and rad i ©response of tumors 
ormal tissue, wi thout extensive reprogramming 
estructuring of the data base, and (3) to pro- 
an interactive, conversational, extensible 
amming environment for both experimentation 
simulation) and clinical application In 
therapy . 



194 



flethods Employed: 

The concepts of decisJon theory are appl 5ed to the 
outcomes of Irradiating a patient, and a figure of 
merit is defined for each treatment plan. This 
figure of merit Involves basic costs, risks, and 
benefits associated with the outcomes, cell survival 
probabilities, and spatio-temporal distributions of 
radiation. Optimization Is equivalent to maximizing 
the expected benefit. The method of solution uti- 
lizes nonlinear mathematical programming with 
linear Inequality constraints. 

flajor Findings: 



e; 
mo( 
c< 
mi 

d . ^ ^ 

computerized bibliography and appropriate search 
program for the investigators research use has been 
devel oped. 

Sign 1 f Icance : 

An analytical approach to radiotherapy optimization 
can mater lal ly improve or, at very least, formally 
validate current rad lotherapeut 1 c practice for the 
following reason: in effect, the analytical approach 
allows the therapist to review al 1 relevant treat- 
ment plans In the light of contemporary rad I ob lol ogy, 
whereas interactive, heuristic approaches now in 
vogue, allow him to review only some possible plans, 
and leave untouched the fundamental Issue of 
criteria for good treatment plans. 




195 



Proposed Course: 

This project vn 1 1 be continued in the coming fiscal 
year, using the PDP-10 computer at DCRT and the PDP-6 
computer at the University of Pennsylvania I-ledlcal 
School. The research plan calls for: (1) explora- 
tion of synthetic treatment problems/ (2) applica- 
tion to treatment plans for actual diseases such as 
cancer of the bladder, (3) extension from two- to 
three-dimensional therapy, and ik) development of 
appropriate CRT displays. 

Honors and Av;ards : None 

Publ icat ions : 

Prewitt, J.S.: Nonlinear Analysis in Biomedical Com- 
puting for Cancer Research. Proc. 1972 Journees 
d ' I nformat ique fledicale, IRIA, Rocquencour t, France. 

Prewitt, J.S.: Programming Systems for 
Optimizing Radiotherapy: Progress and Prospects. 
Invited paper. Second Annual Symposium on the Sharing 
of Computer Programs and Technology in .Nuclear Medicine, 
Oak Ridge IJational Laboratory, Oak Ridge, Tennessee, 1972. 



196 



Serial No. 7.13 

1. Data Management Branch 

2. Office of the Chief, 

DMB 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July I, 1971 through June 30, 1972 

Project Title: A Clustering Algorithm for Biomedical 

Appl icat ions 

Previous Serial Mumber: None 

Project Leader: J. s. Prewitt, DCRT 

Other Professionals: C. Staneck, DCRT 

Cooperating Units: Community Mental Health Center, 

Hahnemann Medical College, 
Philadelphia, Pa. 

Man Years: 



Total : 


0. 


,25 


Profess ional 


: 0. 


,25 


Others : 


0. 


,00 



Project Description 
Object i ves : 



The objectives of this project were (1) to extend 
and refine a hierarchical clustering algorithm de- 
vised by the project leader, (2) to incor- 
porate Calcomp plotter output of program results in 
the form of graphs and taxonomic trees, (3) to apply 
the algorithm to several non-trivial biomedical and 
psychological research problems, and (k) to provide 
a versatile, multi-purpose pattern recognition capa- 
bility for iJ.l.fl. computer users. 



19 7 



Methods Employed: 

Cluster synthesis can be applied to any data base 
consisting of lists of properties or observations - 
discrete or continuous, nominal or quantitative - 
for a sample under study. The program consists of 
modules for: (1) data input, (2) data management, 
(3) executive program control, (.k) assessment of 
inter-object relatedness, (5) execution of the 
cluster merger strategies, (6) construction of taxo- 
nomic trees, (7) assessment of in tra-cl us ter homo- 
geneity relative to inter-cluster heterogeneity at 
any stage, and (8) printed and graphic output of 
intermediate and final results. Evaluation of 
inter-object relatedness can follow any one of 
several methods for computing similarity or proxi- 
mity. Cluster merger strategies can emphasize com- 
pactness, connectedness or central tendancy in the 
resultant groupings. Significance of the spectrum 
or nest of clusters generated by the iterative, 
hierarchical can be judged by inspection of the 
cluster tree v/hich is scaled by a generalized 
within- to total variance ratio. 

Major F indi ngs : 

Two important applications studies were conducted, 
in the first, the repertoire of similarity measures 
and cluster merger strategies v/as extensively exer- 
cised on the (pooled) chromosome arm length measure- 
ments which constitute the Denver-Chicago classifi- 
cation. This widely accepted grouping of chromosomes 
was reconstituted by the clustering program, and a method 
for systematically comparing the results of using the 
various program options was devised. In the 
second application, farthest-neighbor clustering in 
Euclidean space was applied to behavioral profiles 
of 800 elementary school children. The resultant 
behavioral characterization of the clusters was 
significantly correlated with background variables 
and I.Q., and a behavioral taxonomy v/ith predictive 
potential was generated. 



198 



Sign if icance : 

The clustering algorithm Is a valuable and powerful 
heuristic data analysis tool which allows the 
investigator to search for patterns or groups with- 
out Invoking either a priori knowledge or statis- 
tical assumptions about extant categories. 

Proposed Course: 

The program will ultimately be made available via 
the rj.l.H. FORTRAN library to interested users. As 
the need arises, basic capabilities will be ex- 
panded or generalized. In particular: (1) similarity 
measures will be unified in a single formula, (2) 
divisive and nonh ierarch i ca 1 capabilities may be 
incorporated with the current agg lomerat I ve hler- 
archlal procedure, and (3) a classification method 
for assigning unknowns within an established taxonomy 
may be added. 

Honors and Av/ards: None 

Publ i cat ions : 

Prewitt, J. S.: Diagnostic and Predictive Appli- 
cations of Cluster Analysis. Proc. 1971, Journees - 
d ' informat ique fledicale, IRIA, Rocquencourt, France. ^ 

Prewitt, J. f1. S. Experiments with Statistical and * 

Ouas i -Stat i St i cal flethods In Diagnosis. In Jacquez, ^ 

J . (Ed.): Computer Diagnosis and Diagnosis J- 

Methods > Springfield, 111., Charles C. Thomas, 1972. ti' 

(in press) , 

Spivack, G., Sv/Ift, M. and Prewitt, J. M.S.: Syndromes " 

of Disturbed Classroom Behayior: A Behavioral Diag- * 

nostic System for Elementary Schools. J. Special '" 
Education, 1972. (in press) 



199 



Serial No. 7. Ik 

1. Data Management Branch 

2. Scientific Applications 

Sect ion 

3. Bethesda 

PHS-NIH 
individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Cancer Survival System 

Processing of Derma togl yph i c Data 

Previous Serial Number: None 

Project Leader: Thomas J. Heavey, DMB, DCRT 

Other Professionals: None 

Cooperating Units: PSAB, NiCHD 

iian Years: 

Total: 0.20 
Profess I onal : . 19 
Others: 0.01 

Project Description: 

Obj ect i ves : 

The programs developed In this study are directed 
toward providing an objective evaluation of the 
data submitted by Investigators In the field of 
Dermatog 1 yph I cs . The Initial objectives are to 
provide counts of Ulnar and Ridge patterns of the 
fingerprints of about UOl Individuals. The counts 
have been recorded for left and right hands for 
each Individual. Combinations of the counts are 
being made In several ways: by sex, by the pre- 
sence of one when the other is missing, by the 
presence of both, by the absence of both, and by 
combinations between both hands. Upon review and 
study of the first combinations, statistical 
programs will be used to determine what distribu- 
tions exist, or what correlations can be made with 
other factors, or whether there is any indication 
of pathology from any special conditions noted. 



200 



Methods Employed: 

Standard Fortran programming technique has been 
employed, and will be employed in future work, 
except that some procedures may lend themselves 
better to COBOL edit- processing. For any 
Statistical work, the means exist to promptly 
produce the statistics desired. Since the counting 
of the combinations of Ridges and Ulnar patterns is 
somewhat involved, intermediate results were pro- 
duced for a few cases for checking purposes. 

Major Findings: 

Not determined. The initial counts are still 
under study. 

Significance to the Bio-Medical Research Program of the 
I nst i tute: 

Not determined as yet; though the final results 
may not be available for a long time. 

Honors and Awards: None 

Publ icat ions : None 



201 



Serial No. 7.15 

1. Pata Management Pranch 

2. Applied Systems 

Programming Section 
5. Pettiesda 

PHS-NIM 
Individual Project Report 
July 1, 1971 through June 30, 1072 

Project Title: NIH Space 

Previous Serial Number: 7.15 

Project Leader: Felix Liski, PCRT 

Cooperating Units: Space Management Section, ODA, ^, PO 

Man Years : 

Total: n.35 
Profess ional : . 30 
Others: n.05 

Project riescr i pt ion : 

Object i ves : 

The system was revised to make the MjH Space Manage- 
ment Section independent of PCP.T for operating the 
updating and reporting systems. 

The system was revised and streamlined to reduce 
separate submissions of different phases of the update 
system to the computer to one computer run. 

If one phase of the update system detects transaction 
errors, then the subsequent phases will not run until 
the errors are corrected. 

The reporting system was also revised. The PPH pro- 
grams were converted to COBOL language because the 
RPG language is no longer supported. 

The system is being documented and expected to be 
completed by June 30, 1972. 



202 



riethods Employed: 

All input data is captured at a tele-communications 
terminal located in the Space f^anagement Section 
office, and kept in on-line files until used by the 
update segment of the system. (The system combines 
Batch processing, using COBOL and WYLBIIR. ) 

Major Fi nd ings : 

1. Data and submission of computer requests can be 
easily and efficiently entered by non-data pro- 
cessing personnel in Space Management. 

2. Space Management is now about ^5% independent 
of PCRT. 

Sign i f i cance : 

Capture of all input data and submissions of update 
and report requests are done in Space f^anagenent 
on the terminal using the V/YLBUR system. 

Proposed Course: 

For requests of unusual non-scheduled (hut consistent 
in format) reports, one or more Inquiry Reporting 
System (IRS) program packages should be devised for 
use by MSM. 

Honors and Awards: None 

Publications: None 



203 



Serial No. 7.16 

1. Data Management Branch 

2. Software Support 

Sect i on 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1G72 

Project Title: Generated File Reformatting 

Faci 1 i ty ( RE FORf'.ATGEN ) 

Previous Serial fiumber: None 

Project Leader: John R. Parks, Jr., DCRT 

Other Professionals: None 

Cooperating Units: None 

Tion Months: 



Total : 


2.00 


Profess i onal : 


2.00 


Others : 


0.00 



Project Description 
Object i ves : 



Provide a source-program generator such that the 
generated program performs those functions most 
commonly associated with file reformatting. Re- 
duce programmer time to a minimum for a common 
programming requirement. Align the use of this 
function v;ith the existing family of source 
program generators. 



Iiethods Employed 



Using the RflAG language, a source program generator 
(REFORMATGEN) was written to produce a USAS COBOL 
language source program free from compiler errors, 
input requirements and output format v;ere made to 
conform to those for existing source program 
generators . 



204 



Proposed Course: 

Expand the capabilities of REFORMATGEN based on 
user needs. 

Honors and Awards: None 

Publ icat ions : None 



205 



Serial rio. 7.17 

1. Data Management Branch 

2 . Appl ied Systems 

Programming Section 

3. Pethesda 

PHS-N'IH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Framingham Heart Study, Exam 10 

I nven tory 

Previous Serial iJumber: 7.17 

Project Leader: Felix LIski, DCRT 

Other Professionals: None 

Cooperating Units: Biometrics Branch, NHL I 

flan Years: 

Total : n.25 
Profess lonal : 0. 25 
Others: 0.00 

Project Description: 

Object i ves : 

This system is designed to produce descriptive 
frequency tables of symptoms and medical results of 
the tenth set of exhaustive medical examinations 
given to an established cohort of men and women 
living in the area of Framingham, flass. Completed 
reports are evaluated by sponsor, Dev;ey Shurtleff, 
Statistician, Biometrics, iJHLI. 

Methods Employed: 

Input data was Initially keypunched on to nine track 
tapes and edited by the sponsor. Reports were pre- 
pared using COBOL and the Tablemaker Languages. 

Major F Indi ngs : 

N . A . 



206 



Proposed Course: 
None 
Honors and Av/ards : None 
Pub! icat Ions : 

In the near future Section 28 



M L .,, _- of the Framingham Studv 

o^?oidemin!n^'^''/'^'''^'' ^'^^^^ "Frammgham S udy 
of Fpidemiologlcal Investigation of Cardiovascular 



207 



Serial Mo. 7.18 

1. Data Management Pranch 

2 . Appl i erl Systems 

Programming Section 
3. Pethesda 

PHS-NIH 
individual Project Report 
July 1, 1971 through June 30, 1P72 

Project Title: Leukocyte Infusion Update and 

Reporting System 

Previous Serial Number: 7.18 

Project Leader: Vivian Pelham, PCRT 

Other Professionals: G. Pobenecker, DCRT 

navid Rlessley, PCRT 

Cooperating Units: Systems Analysis and Planning Office 

Chemotherapy, NCI 
Leukemia Service, MB, NCI 

Man Years: 

Total: n.50 
Profess ional : 0.35 
Others: 0.15 

Project Description: 

Objectives: 

The purpose of this system is to collect and report 
on Leukocyte Infusion information. The information 
collected includes specific infusion information, 
laboratory follow-up information, and infusion 
series summary information. Information is extracted 
from two related systems (i.e., the Cell Collection 
and the Donor Recipient Information System) and com- 
bined with the Leukocyte information for other 
reports . 



208 



Methods Employed: 

A conversion program was written to make the source 
information as originally coded and punched com- 
patible with the standard update input requirements. 
This was done to give the user the ability to 
insert, modify, and/or delete records using the 
standard updating approach in addition to the 
ability to enter data by coding the original source 
forms. Prior to reporting on the information, a 
program is executed that combines the related re- 
cords of the other systems and creates a combined 
file. The reporting programs address the conbineci 
file, thereby having all the information collected 
and maintained by the other systems available for 
the specific report requirements. A detailed 
Infusion report was written. A summary report of 
infusion series information was written in addition 
to a laboratory follow-up report. 

Major Findings: 

Experiment still In progress. 
Sign i f I cance : 

None ^ 

Proposed Course: ]J 

llhen the system fulfills the collection and re- 
porting requirements of Dr. ("iraw it will be f 
maintained and executed by personnel of NCI. *^ 
Additional programming support will be supplied by i 
the Pata Management Branch of the Division of Com- * 
puter f^.esearch and Technology. * 

l^onors and Awards : None «* 

Publ icat Ions : None 



209 



Serial No. 7. 19 

1. Data Management Branch 

2. Software Support 

Sect i on 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Generated File Reformatting 

Foci 1 ity ( TP.ANSACTGEN) 

Previous Serial Number: None 

Project Leader: John R. Pari<s, Jr., DCRT 

Other Professionals: None oc; 

None 



Cooperat i ng 


On i 


ts: 






lian 1 ion ths : 










Total : 






1. 


,00 


Profess i one 


1 : 


1. 


,00 


Others: 






0. 


,00 



Project Description 
Obj ec t i ves : 



Provide users of DM3 Name Update and Edit Program 
(UPDATEGEN) with a source program generator such 
that the generated program converts files into 
valid input transactions. Minimize the time and 
effort required to implement the DMB Name Opdate 
and Edit Program for users with existing files. 
Expand the capability of the DMB Name Update and 
Edit Program by allowing the output of anoth.er 
system to be used as input. Align the use of this 
function with the existing family of source program 
generators . 



Methods Employed 



Using the RMAG language, a source program generator 
(TRANSACTGEN) was written to produce a U5AS COBOL 
language source program free from compiler errors. 
Input requirements and output format wore made to 
conform to those for existing source program 
generators. 



210 



Proposed Course: 

Expand the capabilities of TRANSACTGEN based on 
user needs. 

Honors and Awards: None 

Publ icat ions : None 



211 



Serial No. 7.2 

1. Data Management Branch 

2. Appl led Systems 

Programming Section 

3. Bethesda 

PHS-IJIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Re-Asslgning Parking Permit flumbers 

when a State Issues fJew License Tags 

Previous Serial iJumber: 7.20 

Project Leader: George Dobenecker, DCRT 

Other Professionals: None 

Cooperating Units: Protection and Parking Section, 

ODA, A, Ps 

Man Years: 

Total: 0.02 
Profess ional : 0.02 
Others: 0.00 

Project Description: 

Object i ves : 



e purpose of the project was to give the Parking 
rm i t Section the ability to update the parking 
rmit master file, by State, when nev; license tag; 
e Issued by that State. 



are 

Method Employed: 

A program was v;rltten to select each employee's record 
from the master file when that employee had at least 
one car registered In a particular State. The pro- 
gram generated a letter to the employee involved. The 
letter shovyed each permit assigned to that employee 
for the State in question and requested that the 
employee enter the new tag number for the vehicle 
bearing the permit(s) listed. The letter also In- 
cluded the necessary keypunch Information so as to 
eliminate the need of transposing the data prior to 
punching. In addition to the letter a mailing label 
and a verification list was generated. 



212 



f'ajor F 1 nd I ngs : 

None 

Sign i f icance : 

None 

Honors and Av/ards: None 

Publ i cat ions : None 



213 



Serial No. 7.21 

1. Data tianagement Branch 

2 . Appl ied Systems 

Programming Section 

3. Bethesda 



PHS-iJiM 
individual Project Report 
July 1, 1971 through June 30, 1972 



Project Title: 
Previous Serial :iumber 
Project Leade r : 
Other Professionals: 



Coope rat i ng Un i ts 



flan Years : 



Total : .".0 

Profess iona 1 : 7.0 
Others: 1.0 



Type 11 Coronary Intervention Study 

7. 21 

Judy ilahaffey, DCRT 

Gary S toner, nCRT 
George Roberts, PORT 
Heorg Schulze, PORT 
George Dobenecker, DCRT 
Roger Dai ley, DCRT 
Claudette f'.i 1 ler, DCRT 
Joe Campbell, DCRT 
Pel ix Li ski, DCRT 
Marty rii 1 ler, DCRT 
Martha fiorton, DCRT 
Dr. John Crensike, CC 

Biometrics Research branch, J ML I 
DCRT, CSL 
DCRT, LAS 
DCRT, DMB, DSSS 

Clinical Center Clinical Programs 
(CCCP) 



214 



Project Description: 

Dhject i ves : 

The MHL I is in the midst of a study to determine if 
lowering the cholesterol level will cause a regression 
or decreased progression of premature coronary artery 
disease. The objective of this project v/as to design 
a complete and comprehensive computer-based system 
for the collection, storage and analysis of this data. 
The scope of this project included the design of data 
collection forms, the development of data entry tech- 
niques, the design of file structures, and the 
development of data retrieval procedures. 

flethods Employed: 

The system encompasses the functions of editing, up- 
dating, monitoring and reporting. It is comprised 
of over 9k programs encompassing four programming 
languages. Standard techniques were employed in the 
areas of input, output, updating and editing. Data 
is being collected via computer-oriented forms de- 
signed especially for this study, laboratory equip- 
ment (analog tapes), and thru interfaces with other 
computer systems that are independent of the Type I! 
system. Correction data is entered by NHL I using 
the CPS facility. The system was designed so that 
it could be operated by personnel in the NHL I . 

Major F i nd i ngs : 

N.A. 
Sign i f i cance : 

N.A. 
Proposed Course: 

The Type II Intervention Computer System is running 
and collecting, storing and editing the data of 
patients. The reporting and post-editing phases of 
the study are now being developed. 



215 



Honors and Av/ards: 

Outstanding performance awards for several individuals 
working on the study and Duality 5^tep increase for the 
Project Leader. 



Publ icat ions 



None 



216 



Serial No. 7.?? 

1. Data flanagement branch 

2 . Appi led Systems 

Programming Section 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Railroad Retirement 

Previous Serial Number: 7.22 

Project Leader: Judy Mahaffey, nCRT 

Other Professionals: Penny Brogan, PCRT 

Pel ix L i ski , nCRT 
Charles Twigg, DCRT 
Vivian Pelham, DCRT 

Cooperating Units: National Cancer Institute 

Man Years: 

Total: 1.27 
Profess ional : 0.97 
Others: 0.30 

Project Description: 

Objectives : 

The objective of this project is to make it possible 
to do statistical processing on the Railroad Retire- 
ment data by creating a single, standard data file 
from the existing incongruent files which have 
evolved over the past ten years. This requires a 
complete analysis and redesign development of pro- 
grams to convert the existing files and complete 
documen tat ion . 



217 



Methods Employed: 

The Railroad Retirement data files, consisting of 
over two million records in all, have evolved over 
the past years and have resulted In 7-P different 
data formats and coding structures. In the systems 
analysis phase, all existing coding structures and 
formats were completely documented and correlated. 
A standard format, common coding scheme, and data 
validation method was developed for all files to 
be converted. A series of COBOL programs were 
executed to accomplish these tasks. Check digit 
procedures and control records were established to 
insure accuracy. 

Sign i f i cance : ' 

The effort described above will make it possible 
for MCI to do a meaningful statistical analysis of 
the Railroad Retirement data. This design will 
facilitate analysis by tailoring processing to be 
compatible with the PCRT library programs. 

Proposed Course: 

MCI has a standard data file and is conducting 

statistical analysis on the Railroad Retirement 

data. Documentation is to be completed and turned 
over to MC I . 

Honors and Awards: None 

Publ icat ions : None 



218 



Serial rJo. 7.2 3 

1. Data rianagement Rranch 

2. Appl led Systems 

Programming Section 

3. Bethesda 

PHS-iJIH 
individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: URBS Systems Study 

Previous Serial number: 7.25 

Project Leader: Judy Mahaffey, DCRT 

Other Professionals: None 

Cooperating Units: Unit for Research into Behavioral 

Systems, f! I MH 



Man Years: 






Total : 


0. 


,08 


Profess ional : 


0. 


,08 


Others : 


0. 


,00 



Project Description: 

Object i ves : 

The purpose of this project was to aid the Section on 
Behavioral Systems, iJMH in establishing a computer 
system for statistical system to do statistical 
analysis of the Population Study data they have 
collected over the past few years. 

f-ethods Hmployed: 

A systems-approach to data analysis and file design 
was presented to URBS personnel. These techniques 
were implemented by URBS personnel in 1) defining 
their information goals, 2) evaluating their current 
data collection procedures and 3) defining and docu- 
menting the current status of their data. Guidance 
was provided through these stages. A computer-based 
data storage system was proposed. This included file 
structures and data coding techniques. f'ethods for 
converting existing data to the computer data base 
were recommended. Guidelines for documentation, 
control procedures and future data collection were 
presented. The project is not complete. 



219 



Serial No. 7.?U 

1. nata Management Pranch 

2 . Appl ied Systems 

Programming Section 

3. Rethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 3n, 1072 

Project Title: Histocompatibility Computer Analysis 

Previous Serial Number: 7.24 

Project Leader: Claudette Miller, PCRT 

Other Professionals: None 

Cooperating Units: Leui<emia Service, M.P., NCI 

Systems Analysis and Planning 
Office, Chemotherapy, NCI 

Man Years: 

Total: n.50 
Profess ional : n. U8 
Others: 0.02 2 

Project Description: ^ 

Objectives: •■ 

The objectives of this design ^r& to allow for com- f^ 
puter update and correction of data, to facilitate 
data retrieval to provide for analysis in evaluating 
test results, to mechanize the conversion of the 
raw data to the present data base, and to educate 
personnel in the use of the system. 

Methods Employed: 

Standard Information Analysis techniques have been 
used. Programs were written to eliminate the time- 
consuming manual v/orkload, while analysis of the 
system was being done. The initial system has been 
expanded and supporting programs are now being 
wr i tten. 



221 



significance to Plo-medical Research and the Program of 
the I nst i tute : 

This system is intended to aid in accurate HL-A 
typing, which is necessary to determine acceptable 
white cell and platelet transfusions and bone marrow 
transplants for leukemia patients. It is believed 
that the establishment of files will allow more 
sophisticated analysis of the experimental results. 

Proposed Course: 

Implementation of the system will be completed by the 
end of this fiscal year. A request has been made for 
additional programming to alleviate the manual work- 
load. This programming v/ill be done v/h i 1 e implemen- 
ting the system. There are plans for investigating 
the initial analysis of raw data to determine the 
feasibility of computer analysis. 

Honors and Awards: flone 

Publ i cat ions : None 



222 



Serial No. 7.25 

1. Data Management Branch 

2. Software Support 

Section 
5. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Generated CPS Prompt Program 

Previous Serial Number: None 

Project Leader: Darius Georg, DCRT 

Other Professionals: Robert A. Magnuson, DCRT 

Cooperating Units: None 

Man Years: 

Total: 0.30 
Profess iona 1:0.30 
Others: 0.00 

Project Description: 

Obj ect i ves : 

Provide a CPS program v/hich is a tool to be used 
as an aid to collect transaction records to be 
used as input to the name update and edit system. 
The program is designed to prompt the user for 
predefined input thus reducing the possibility of 
error during collection of the raw data. 

Methods Employed: 

An RMAG language program is used to generate a CPS 
prompt program. The RMAG program is written to 
accept the same macro calls used for the genera- 
tion of the update and edit program. The CPS pro- 
gram v>/ill now prompt the use for the same field 
names that are known to the update and edit 
program. 



223 



The following is a list of features of this module: 



1. 
2. 



Prompts for input by field name 

Accept only numeric input when the field was 

defined as a numeric field. 

Reject the input if it is longer than the 

field 1 ength . 

Provides editing for acceptable values. 

Provides editing of data being entered on the 

terminal with inbedded decimal points. 

Builds transaction records in a free form that 

is acceptable as input to the update and edit 

program. 



Proposed Course: 

The system may be extended for a larger set of 
requirements allowing for more editing logic and 
to allow the output to be varied based on the 
users specifications. 



Honors and Awards 



None 



Publ i cat ions 



None 



224 



Serial No. 7.?f; 

1. Pata Management ''ranch 

2. Software Support Section 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1^ 1971 through June 50, 1P72 

Project Title: Emergency Virus Isolation Facility 

Medical Monitoring Project 

Previous Serial Number: 7.26 

Project Leader: John Parks, nCRT 

Cooperating Units: Biohazards Containment and Control 

Section, VO, FT, MCI 
Employee Health Service "ranch, CC 
Computer Center "ranch, PCRT 

Man Months: 

Total: 3.0 

Profess ionsl : 1.0 
Others: 2.0 

Project Description: 

Object i ves : 

This system is designed to convert FV I F employee medi- 
cal records to a computer-based file system and to 
provide basic maintenance and reporting capabilities 
for the file. 



225 



Methods Employed: 

The DMB Standardized Update System was used for file 
creation and maintenance. Input is typed into a 
IVYLBUR dataset from special pre-printed forms. The 
system is operated entirely via UYLBUR with input 
data, rejected data, and run JCL on l.'YLBUR datasets. 
Ruilt into the file maintenance procedure is a step 
which maintains volume rotation for all JCL datasets. 
This feature reduces operation of the system to merely 
•USE'ing and 'RUN'ning the desired UYLBUR JCL dataset. 

f'ajor F ind i ngs : 

Although the system is operational, personnel shortage 
prevented the user from accepting operating responsi- 
bility. Instead, the system vias documented to PCRT 
Standards and submitted to the Systems Application 
Section as a recurring production system. 

After approximately one year of operation, experience 
and changing demands have brought about the necessity 
for a major revision of the coding forms and related 
programming changes. This development temporarily 
suspended efforts to convert the system to a recurring 
production status. Also brought out was the fact that 
changes of this magnitude can be expected at least 
every two years v/ith each new Public Health Service 
Doctor assigned to the project. 

Proposed Course: 

The coding forms have been redesigned, approved, and 
submitted for printing. The necessary programming 
changes have been made along with conversions from 
F-LEVEL to USAS COBOL of all programs. Some coordi- 
nating will be necessary to achieve a smooth transi- 
tion to the new coding forms along with some 
instruction for concerned personnel during the transi- 
tion period. Conversion of the system to a re- 
curring production status v/ith operation and maintenance 
by the Systems Application Section to be accomplished 
by the end of the fiscal year. 

I^onors and Awards: None 

Publ i cat ions : None 



226 



Serial tJo. 7.27 

1. Data Management '3ranch 

2. Software Support 

Sect Ion 

3. Bothesda 



PHS-MIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 



Ouick Report Oenerator 

None 

Robert A. nagnuson, DCRT 

None 



Proj ect Title: 

Previous Serial Number 

Project Leader: 

Other Professionals: 

Man Years: 

Total: 0.15 
Profess i onal : n. 15 
Others: 0.0 

Project Description: 

Obj ect i ves : 

Provides a powerful report generator for c;bst ract i n? 
information from disk files or tapes. it is note- 
worthy for its comprehensive logical capability. 

flethods Employed: 

An RMAG program was designed. Implemented, and 
tested. From easy to prepare input, this program 
generates IRS statements which actually interrogate 
the files to produce the desired report. 

Major F I nd i ngs : 

The use of the program indicated that this nethod, 
i.e., RMAG-to- I RS- to-report , is cost effective. 



227 



Si gn i f i cance : 

This program is one of DMB's RMAG-coded source- 
language generators. The data-selection logic can 
be complicated. It can use 5 different boolean 
connect I ve--negat I on, conjunction, alternation, 
conditional, and biconditional. Parentheses are 
allowed to almost any level of nesting. Since the 
target language (IRS) allows only 2 connectives and 
no parentheses, an important subprogram transforms 
the user-supplied boolean expression into (dis- 
junctive) normal form (from which It Is further 
transformed into the actual IRS select statements). 

Proposed Course: ^ ^ _^ 

The system will be extended and more broadly applied 

Honors and Awards: None 

Publications: None 



228 



Proj ect Title: 



Serial No. 7.28 

1. Data Management Branch 

2. nocumrnta t i on f^ Systems 

Support Section 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 tbrou?:h June 30, 1972 

Trends in Graduate Enrollment and 
Ph.D. Output in selected Science 
and Health Professional Fields. 



Previous Serial Number: 
Project Leader: 
Other Professional: 

Cooperating?; Units: 
flan Years: 



None 

Charles Twigg 

D. George 
T.J. Heavey 

OPPE/OP, NIH 







Total : 


0. 


,50 






Profess i onal : 


fj 


,50 






Others: 


0. 


,00 


P 


roj 


ect Descr i pt i on ; 
Obj ecti ves : 










1. Create 


data 



Create data base, from input data from Office 
of Education and NIH, which will be compatible 
for editing, updating and report processin-^. 

Produce tables for publication by Manpower 
A f! a lysis Rranch. 

Provide capability for statistical analysis and 
tota 1 i ng. 



229 



flethods Employed: 

1. Programs written to edit and balance all data 
rece 1 ved . 

2. After data was edited and balanced, desirned a 
record layout for all data for all years. 

3. Established procedure to add new data each year 
and delete the oldest year since this data base 
will always contain the latest 10 years for 
reportinp: nurposes. 



Major F I nd i nss : 

Fstabl i sh i n;' a compatible data base for all input 
and carryint^ it forward year to year will save edlt- 
inr and balancin.e time for all r.revious years' date. 
Ip addition all personnel will be workin? with the 
the same standard record format v/hich will allevi.te 
most of the communication problems that have arisen 
in the past between the different f^roups workinr 
on this proj ect . 

: ) i g n i f i c a n c e : 

.None 
Proposed Course: 

Giipport future years of this project as reouired. 
Honors and Awards: None 
Publ i cat i ons : 

Tables by Health Manpower Branch, on/NIH 



230 



Serial No. 7.29 

1. Data Management Branch 

2. Scientific Applications 

Sect i on 

3. Bethesda 



PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Cancer Survival System 

Previous Serial Number: 7.29 

Project Leader(s): Catherine Staneck, DCRT 

Other Professionals: Frances Bauer, DCRT 

Lee McBride, DCRT 

Cooperating Units: End Results Section, 

Biometry Branch, NCI 

Iian Years: 



Total : 


0.56 


Profess ional : 


0.56 


Others: 


0.00 



Project Description; 
Obj ect i ves : 



The Survival System was originally developed seven 
years ago to support the End Results Section, NCI, 
in preparation of the End Resul ts in Cancer 
Reports. Maintenance and improvement of that 
system is the primary goal of the project. 



Methods Employed 



The system was modified to add computation of 
median survival time for each data classification 
within a g i ven run. 

An important extension to the system was inclusion 
of an optional point plot routine which was in- 
tegrated into the system in such a way as to allow 
plot information to be obtained directly from other 
system modules. This routine may be invoked by the 
inclusion of appropriate control statements in the 
same manner as the RECODE, SORT, COUNT, and OUTPUT 
modules are invoked. 



231 



Major Findings: 



The SUfi program was originally designed for maximum 
utilization of core but this design was not 
implemented until this year. Implementation re- 
sulted in a considerable savings in run time. 



S i gn I f i cance : 

The Survival System or modifications thereof have 
been utilized for a number of other projects in- 
cluding a study of survival rates for heart valve 
replacements and replacement techniques and the 
Cigarette Condensate study. 

Copies of the system have been sent to the Boston 
University School of fledlcine and the Tumor 
Registry at Connecticut State Department of Health 

Proposed Course: 

Continuing maintenance of the system. 

Honors and Awards: None 



Pub 1 i ca t i ons 



Unknown 



232 



Serial Mo. T.'^O 

1. Pata Management Pranch 

2 . Appl i ed Systems 

Programming Section 
3. Pethesda 

PHS-NIH 
individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: NIH International Activities 

Previous Serial Number: 7.30 

Project Leader: Roger nailey 

Other Professionals: None 

Cooperating "nits: Fogarty international Center 

Man Years: 

Total: n.20 
Profess lonal : 0. 12 
Others: 0.08 

Project Oescription: ■*] 

Object I ves : «* 

The purpose of this project is to design a system ^or *« 

the storage, maintenance and retrieval of information L 

on NIH International Activities. All transactions on i;?': 

foreign research and training grants and contracts are •» 

now being maintained manually. The system, when com- gl 

plete, will be capable of maintenance and operation J 

by the Fogarty Center. "*' 

Methods Employed: 

The system is still in the design phase. f'ecause many 
of the data elements needed to establish the Inter- 
national Activities data base currently exist in the 
nRG computer files, the design of the system will in- 
corporate a method of extracting that data. A main- 
tenance and reporting system is now being developed 
to meet the Fogarty Center's U8-hour demand 
requ I remen ts . 



233 



Major Fi nd ings : 

None 
Sign i f I cance : 

None 

Proposed Course: 

V/hen the design phase has been completed and the 
system is fully operational^ the complete system will 
be handed over to the Fogarty Institute. The system 
will be turned over to the Fogarty Center in stages, 
as each stage becomes operational. 

^'onors and Awards: None 

Publications: None 



234 



Serial No. 7.31 

1. Data Management Branch 

2. Applied Systems 

Programming Section 

3. Rethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Medical Records: Diagnosis 

and Discharge 

Previous Serial dumber: 7.31 

Project Leader: Penny Brogan, DCRT 

Other Professionals: David Blessley 

Cooperating Units: Medical Records Department, CC 

Man Years: 

Total : 1.2 
Profess ional : 1 . 2 
Others: 0.0 

Project Description: ^ 

MM 

MM 

Objectives: ^ 

The purpose of this project is to design and imple- f 

ment a system for maintaining a computer file with f|= 

basic diagnosis Information on NIH patients. The file ' :- 

Is composed of identifying information on each patient *" 

and the diagnosis rendered each time the patient Is "«| 

discharged. Various selective listings and statistics gf 

will be produced from the data. ,«i 

flethods Employed: 

Data will be collected and corrected on UYLBUR and 
stored on 9-track tape. COBOL programs have been 
written to update the file and produce desired reports. 



235 



S ign i f icance : 

The system will help the meHical records librarians 
bring all their diagnostic data up-to-date. '/hen the 
system is computerized it will be easier for the 
librarians to maintain the accuracy of the diagnostic 
records and to obtain composite lists as needed. 

Proposed Course: 

The system has been defined and programmed. Final 
JCL for each phase is now being set up and documen- 
tation is being completed. 

Honors and Av/ards: None > ^ "^^ «=, 
Publications: None ■ ' 



236 



Serial No. 7.32 

1. Data Management Branch 

2. Documentation ^j Systems 

Support Section 

3. Bethesda 

PHS-MiH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Hepatitis Australian Antigen File 

Study. 

Project Leader: Barry Madia 

Other Professions: Louise Saulnler 

Georg Schulze 

Cooperating Units: Clinical Center OCAMS 

T'lan Years: 

Total: 1.50 
Prof essi ona 1 : 1.50 
Others: n.OO 

Project Description: 

Obj ect i ves : 

The study consisted of creation of a patient naster- 
file of test and control subjects selected from the 
NIH personnel file. The test population consisted 
of physicians, dentists and dental technicians, 
nurses and blood bank personnel - people havinr the 
opportunity to handle blood and blood products in 
their work environment. The control population was 
selected from NLM, DCRT, and ORG personnel. The 
control and test population were further sub- 
divided by classifications as to race, GS level or 
PUS rank, sex and age. A set number of patients 
had to be selected and tested from each of the 
groups defined by the classes named above for 
the analyses phase of the project. Testing con- 
sisted of completion of a questionnaire ,<rivinr 
identifying^ personnel data and a history o-p 
exposure to hepatitis. This information was 
recorded on the patient masterfile. A lab result 
file was also maintained which recorded the lab 
results from 3 different lab tests performed on 
two vials of blood drawn from each subject. 

237 



Methods Employed: 

Subject data was recorded on a form completed by 
a testee at the time when the two vials of blood 
were drawn. The vial numbers of the tubes drawn 
were recorded on the patient questionnaire at the 
time of testing. The transaction files for the 
patient master and lab result master were created 
by CPS prompt routines. The patient master update 
and several utility routines for detecting the 
status of data collection were written in the COBOL 
language. The lab result update and update reports 
were written in PL/1. After completion of pro- 
gramming the system was turned over to clinical 
center personnel for continuing data collection and 
running of the updates. 

Major Findings: 

Experiment still In progress. 
Si gn t f icance: 

None 

Proposed Course: 

Data collection has been going on since July 1971 
and is nearing completion. Additional programming 
support will be needed for the analysis phase and 
will be supplied by DCRT. 

Honors and Awards: None 

Publications: None 



238 



Serial Mo. 7.35 

1. Data Manap-ement Branch 

2. Documentation ^ Systems 

Support Section 

3. Bethesda 



PHS-MIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 



Proj ect Title: 
Project Leader: 

Other Professionals 



Cooperat i np Uni ts 



Case Reports 

Loui se Saul ni er 
Charles Jvj'\ gp; 

Georg Schulze 
Pel i X Li ski 
George Dobenecker 
Dennis George 

Office nf Resources Analysis, 
OADPPE, Applied Systems Prorrrimrni nf; 
Section, DMB 



Man Years: 




Total : 


1.20 


Professional : 


: 1.20 


Others: 


0.00 



Project Description: 
Obj ect i ves : 



The project consisted of three phases: DCol lection 
of data from all reporting DHEW agencies of money 
given in grants for research and development to 
institutes of higher education, health professional 
schools, non profit organizations and hospitals 
2) A submission of data and an accompanying roport 
to the National Science Foundation 3) Generation of 
the CASE reports for Department of Health Education 
and V.'elfare five volume publication. 



239 



Method Employed: 

The CASE reports are a recurring yearly project which 
runs approximately nine months. This year the 
specifications for reporting of data were redefined 
and an additional card format was designed for break- 
ing down money reported for research and development 
to institutions of higher education by field of 
science. Programs written for maintaining the data 
masterfile, balancing money reported by each agency 
with accompanying tables, and editing data before 
updating the masterfile were rewritten because of 
the new specifications. The submission for National 
Science Foundation was in two parts: higher education 
and non profit organizations. After all data had 
been submitted the data base for the generated 
reports was created. Several new tables were 
requested for the publications this year requiring 
additional programming effort. 

Major Findings: None 

Significance: None 

Proposed Course: 

New specifications for data submission for next 
year are in the process of being written. New re- 
porting specifications and a revised data record 
format will require modifications to all programs 
written for data collection. The project next year 
will run much the same as in past years. 

Honors and Awards: None ' 

Publications: None 



240 



Serial No. 7.3tt 

1. Data Manap;pment f3r?nch 

2. Documentation <" Systems 

Support Section 

3. Pri-hesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Nursing Education Study 

Previous Serial Numbers: None 

Project Lender: Mr. Thomas J. Hoavey, nMB--nCRT 

Other investigators: ^lone 

Cooperating Litilts: Nursing Educ. Office, Clinical 

Center, NIH Hath. Stat. Section, 

DHB 

f'an Years: 



Total : 


0.36 


Profess i onal 


: (J. 33 


Other: 


0.03 



Project Description 
Obj ect i ves : 



To depict with simple statistics (means, S.P's, 
T-Tests) the results of 5 years of testing on the 
performance of about 250 Nurses of two main hack- 
grounds--one group from r^acca 1 aureate Degree grant- 
ing institutions and one group from Medical Tech- 
nology Schools with three-year Certi^^icate Programs. 
The exposure of the two groups to various training 
and experience situations during their tour of duty 
at the Clinical Center was also investigated and the 
same simple statistics computed for Ihk items of 
experience, taken at three time intervals. Thus 
two phases of the study were undertaken, with some 
changing of the individuals observed (duo to 
necessary rotation of their duties), in order to 
gain some insigh.t and make recommendations about 
future Education and Training pro'^rams for '''urses to 
be applied to other locations in adrlition to the l^W^. 
CI i n i ca 1 Center. 



241 



Methods Employed: 

Standard IBM Fortran IV Programming techniques were 
employed. Pre-programmed procedures were not used 
as the number of variables, the special formats for 
output/ and the necessity to eliminate missing data 
and devise self-editing to explain why certain cases 
would not work. The variable sample sizes, and large 
requirements for memory use precluded using pre- 
programmed procedures without extensive modification. 
Correlation coefficients were produced for those 
cases in which sample sizes were equal, means and 
S.D.'s were given except where data was missing 
entirely, and degrees of freedom were computed for 
all valid cases using B. Ostle's 3rd hypothesis* 
for T-Tests. 

* Ref: Bernard Ostle, "Statistics in the Physical 
Sciences". 

A Function of the form R=F (A-(BxC))/D) was used re- 
peatedly to shorten coding in the computational sub- 
routine developed. 

Major F i nd i ng : 

1. No analysis of results is available, but the 
T-Tests and other statistics are apparently 
acceptable to the customer Institute. 

2. Useful means were developed for computing T- 
Tests and simple statistics without using 
large-scale statistical routines, and with- 
out presenting the customer with more in- 
volved results than desi red-- i . e. , the 
customer was presented with a tailor-made job 
which not only produces the results required 
in the formats desired, but also analyzes 

the input selected for comparisons and de- 
scribes why certain cases cannot be computed. 
That is, the procedure is, to a large 
extent, self-editing. 



242 



Proposed Course: 

Unknown, but apparently results will allow user 
to decide future policy in regard to the selection, 
training, and exposure to experience-situations for 
Nurses in the NIH Clinical Center and elsewhere. 

Honors and Awards: None 

Publ i cat ions : 

The output of the series of programs will be used 
as backup for a Study to be published by Dr. 
Margaret Dunn PHD, summarizing and recommending 
improvements in the Nursing Education Program. 



243 



Serial No. 7.?5 

1. Data Management Branch 

2. Documentation ^ Systems 

Support Section 

3. Bethosda 

PHS-MIH 
IndividLial Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Data Retrieval from the Heta File 

Previous Srrial fJumbers: None 

^Toject Leader: Dennis Oeorf-e 

Other Professionals: 



'" a r r y f 1 a d i a 
Georjr Schul ze 

NIAMD, NHLi, NIAID, f'CI, C.C. 



to provid'-^ clinical 
patient Laboratory 
Center Heta System 



Cooperating Lip its: 

flan Years: 

Total: 0.50 
■^ro^ess i ona 1 : 0. fiO 

Others: 0.10 

Project f'tscr i pt i on : 
Oh i ect I ves : 



The purpose of this '^reject is 
Investigators with surfiles of 
Test Results from the Clinical 
request . 

"g thods Fnipl oyed : 



The methods employed are as varied as the renuests 
they must satisfy. S^^veral generalized COSOL ^nd 
PL/1 routines have been developed v/'ich select the 
desired data and produce reports to satisfy the 
majority of requests. 

ij or F I nd i ncs : 

The system has proven that 1) data not previously 
available or available only after exhaustive clerical 
searchin,^ thru medical records can be obtained 
quickly and t-asily 2) there is an increasin need 
for the type of data available on the -^ i 1 e . 



244 



Proposed course: 

The future plans are to continue development of 
standardized routines to encompass more types of 
requests and possibly to rearrange the files to 
reduce cost of retrieving data. 

Honors and Awards: None 

Publications: None 



finit 



245 



Serial No. 7.56 

1. Data Management Branch 

2. Documentation ?>. System 
Support Section 

3. Bethesda 

PHS-K'IH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Vitamin C Study 

Previous Serial Numbers: None 

Project Lender: Dennis Heorge 

Other Prof es s I onr, 1 s : r'arius Georg 

Cooperating Units: C.C., NIH 

Software Support Section, OflB 

•'■'Dn Years: 

Total: 0.30 
Prof essi onTil : 0.10 
Other: D . 2 



Project Description: 
Ohj ect I ves : 



The purpose of this project Is to design and imple- 
ment a system to 1) collect ^ata on people enrolled 
in the 'Vitamin C' study 2) edit the data and create 
files which will be used in the statistical analysis 
3) provide the statistical analysis required to 
determine the effect of Vitamin C on the common cold, 

[iethods Employed: 

A CPS prompt system was designed and Implemented for 
collecting and editing the data. The data ar'^ 
collected and if they pass the edit, are formatted 
as transactions for the standard update system. Tf^e 
files are then updated using standard update. 



246 



Major Findings: 

/Although) the data is only partially collected at this 
time the CPS prompt routine has proven an effrctive 
method for collecting data. It provides both data 
security and an immediate edit, no invalid data is 
accepted onto the system. 

S i gn i f i cance : 

Same as Major Findings. 

r'roposed Course: 

Data collection will continue until all data have 
been collected. The statistical analysis will then 
be completed to determine the effects of Vitamin C 
on both common cold prevention and duration. 

Honors and Av/ards: None 

Publ icat ion: None 



247 



Serial No. 7.57 

1. Data Management Branch 

2. Documentation and 

Support Section 

3. BethesHa 



PHS-NIH 
Individual Project Report 
July 1, 1971 throu?:"h June 30, 1972 

Project Title: Documentation Standards 

Previous Serial Number: None 

Project Leader: John A. Haggerty, Jr. 

Other Professional: Chief, PilB 

M. Vincent, DMB 
Section Head, DMB 

Cooperating Units: All Sections, DMB 

.'Ian Years: 

Total: 0.30 
Profess I ona 1 : 0.30 
Others: 0.00 

Project Description: 

Ob j ect i ve : 

To establish an outline, for use of Systems Analysts 
and Programmers in the Data Management Branch, that 
will control standards and assure that each system 
and/or program can be operated according to each 
set of documentation guidelines by the user. 

Methods Employed: 

A check list of Items was initially drawn up for 
both system and program documentation, presented 
to all section heads of DMB with requests for 
additions, deletions and comments from them. l.'hen 
all section heads agreed on the items to be Includ- 
ed, the outlines were drawn up and were adopted by 
the branch. 



248 



Major Findings: 

The documentation outline provides the prop-ranmor 
and analyst with a convenient checklist for assuming; 
that all necessary documentation is prepared and 
since its format is standardized, references to 
each set of documentation are facilitated. 

S i gn i f i cance : 

All documentation will be standardized so that any 
one looking for a particular segment will be able 
to find it easily. It will also allow for personnel 
other than those who wrote the original documenta- 
tion, to follow the operation of the system and/or 
system and/or program. 

Proposed Course: 

To provide a follovj up procedure for insuring that 
all sections of DMB are writing their documentation 
according to the outlines established. To update 
outline as deemed necessary. 

Honors and Av/ards: None 

Pub 1 i cat i ons : 

These documentation standards were followed in 
the new Users Guide published by the Math/Stat. 
Section of the Data Management Rranch. 



idjj) 



«aElt 



249 



Serial No. 7.58 

1. Data Management Branch 

2. Documentation and 

Support Section 

3. Bethesda 



PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 



Proj ect Ti tie: 



N 



CHD Patient Maintenance 
& Retrieval System 



Previous Serial Number: 

Project Leader: 

Other Professionals: 

Cooperating Units: 

Man Years: 

Total O.kO 
Profess ional : 0. UO 
Others: 0.00 

Project Description: 



None 

Ba r ry Mad i a 
None 
Intramural Research OADP NICHD 



Obj ect i ves 



The 
tr i e 
out- 
col 1 
k i n 
phys 
date 
di ag 
ta i n 
requ 
ages 
were 
to t 
up d 



purpos 
ve sel 
pat i en 
ected 
res i de 
I c i an 

and d 
nos i s 
i ng to 
es ted 
, sex, 

w r i 1 1 
he ref 
ata to 



e of 
ect i 
ts a 
pert 
nee, 
at t 
iagn 
code 

dec 
to 1 

and 
en t 
er r i 

kee 



the 
ve i 
dmi t 
a i ne 

ref 
he N 
OS i s 
s, f 
ease 
i St 

d i a 
o cr 
ng p 
p NI 



system wa 
nf orma t i on 
ted by the 
d to patie 
err i ng phy 
I H, nurs i n 

codes, di 
ol 1 ow up d 
d pa t i ents 
pa t i en ts o 
gnos i s cod 
eate corre 
hys i c i ans 
CHD record 



s to col 
on i npa 
NICHD. 
nt res i d 
s i c i an, 
g unit, 
scharge 
ate, and 
Retri 
f spec i f 
es. Spe 
spondenc 
or pat i e 
s curren 



lect and re- 
t i en ts and 

The data 
ence, next of 
at tend i ng 
admi 1 1 i ng 
date and 

data per- 
evals were 
ic doctors, 
cial programs 
e to be sent 
nts for f ol low- 
t. 



250 



Methods Employed: 

A set of CPS programs was written to collect data 
for the DMB generated name Update System which was 
used. The data was validated at time of collection 
and transactions were written to an on-line data 
set. During; the data collection, the user has the 
option to have an update run scheduled throupih CPS 
R JE . 

A special CPS prompt program was designed to prompt 
the user for data items to be retrieved, printed, and 
sorted, and to include any report headings desired. 
These collected items v/ere passed to a Rf'AG pro- 
gram that created IRS statements, which in turn 
were used to retrieve the desired data and store 
it on an on-line data set. This data set was then 
accessed with a CPS program for review. 

The special correspondence programs were written 
using COBOL. These programs selected patient 
data necessary for correspondence by matching 
patient numbers, or by finding a specific -PoIIow- 
up date in the records. All correspondence is 
written on a WYLBUR data set then printed on NIH 
1 etterhoad. 

S i gn i f i cance: 

The MICHD is now able to keep current data on all 
of their patients thru their correspondence 
programs and data received from the medical records 
department . 

in the past all data was kept on index cards and 
finding information was very time consuming. 

The correspondence programs have saved the 
secretary considerable time in typing. 

Proposed Course: 

Additional support will be supplied upon request. 
Honors and Awards: None 
Publ icat ions : Mone 



251 



Serial No. 7.39 

1. Data Mana<?:ement Branch 

2. Documentation and 

Support Section 

3. '^othesda 



Proj ect Ti tie: 



PUS-NIH 
Individual Project Report 
July 1, 1971 tbrour.h June 30, 1972 

C.C. Patient Rep;ister 
and Plate 11 a k i n "; 



Previous Serial Number: 

Project Leader: 

Other Profess ional : 

Cooperat i n,'!: l.inlts: 

■Ian Years: 

Total: n.i+0 
Profess! onal : 0. itO 
Others: 0.00 

Project Description: 

Objectives: 



None -^ ^ ?^= 
R. [lad! a 
0. Schu 1 ze 
OCAMS, C.C. 



1. Calculate check oi,?^it and enter Into last 
position of patient number oxpandln?^ field 
f rem 6 to 7 digits. 

2. List patient re^^Istry file. 

3. Append additionr.l data to re,'^istry file. 

k. Create tape and punc'-'ed cards to he Lised 
for addresso.f^raph plate makln.n:. 



252 



Method Employed 

A program was written using the specifications 
supplied by OCAMS, C.C. in "Check-Digit Technical 
memorandum" February 1, 1972 and the patient 
registry list was run from a tape supplied by the 
admissions system of the Clinical Center with the 
new 7 digit number. Appending the additional data 
involved making use of the Clinical Patholo<^y and 
niagnosis Discharge Files and Discrepancies or 
missing data between files were flagged for updat<^ 
and correction by C.C. personnel. 

Major Findings 

There were numerous instances of discrepancies or 
missing data either in one, two or all of the 
f i 1 es. 

31 gn i f icance: 

Creation of the Patient Registry List will enable 
personnel in the C.C. to locate patients usinr 
one combined source. 

Proposed Course: i«**; 

Files will have to be updated and corrected before '"*| 

addressograph plates can be used effectively. "'*' 

After this tape Is created the C.C. con request if" 

listings or statistics of any or all data elements ■-- 

in the file (e.g., list by patient name, by date ir' 

of birth, etc.). The addressograph plates will be «") 

used for necessary mailing. «?' 

..If 
!!onor and Awards: None 



Publ I cat ions : None 



mm 



HHI^t 



253 



Serial Mo . 7 .kO 

1. Hata Management branch 

2. Mathematical and 

Statistical Section 
^. Pethesda 



PHS-NIH 
Individual Project Report 
July 1, 1971 through June 50, 1072 

Project Title: Biologies Controls Systems 

Previous Serial Mumber: None 

Project Leader: Fred Yamada, DCRT 

Other Professionals: None 

Cooperating I'nlts: Biometrics Section, HBS 

Man Years: 



Total : 


n, 


.30 


Profess ional 


: 0, 


.30 


Others : 


n, 


.no 



Project Description 
Object i ves : 



The programs developed in this study are directed 
toward providing an objective evaluation of the 
data submitted by manufacturers in support of 
license applications and/or release of production 
lots or data developed by the staff of PBS in their 
search for improved methods of assessment of 
biologies products. 



254 



Work in FY '72: 

Ploassay programs were developed to compute (a) 
either problt or logits, (H) either individual or 
pooled slopes, (c) v/ith or without conversion of 
dose input, (d) either initial values to be cal- 
culated or read in, (e) both a test for parallelism 
and relative potency to be calculated. 

The other major task In this area is the develop- 
ment of a trichotomous assay program which, when 
available, will be applied to an extensive 
accumulation of data. 

Major Findings: 

Not appl i cable 

Proposed Course: 

Complete this task v/ith further Improvement, de- 
bugging and generalization of the system to improve 
methods of assessment. 

Honors and Av/ards: None 

Publications: Pending ia«> 



0ai 
(iiinl' 



255 



Serial No. l.kl 

1. Hata fianasenent Pranch 

2. Matheniat i cal and 

Statistical Section 

3. PethesHa 

PHS-NIH 
Individual Project Report 
July 1, 1P71 through June 30, 1Q72 

Project Title: V/ard Value Orientation Study 

Previous Serial Number: None 

Project Leader: Dave VanSant, PCRT 

Other Professionals: None 



Cooperating Units 



Experimental Oroup and Pamily 
Studies Section, Adult Psychiatry 
Pranch, NIMH 



Man 


Years: 








Total : 


n. 


,30 




Profess lonal 


: n, 


,30 




Other: 


n. 


,no 



Project Description 
Packground : 



A questionnaire consisting of ahout 150 response 
items is being administered to certain doctors, 
nurses, and psychiatric patients in several Insti- 
tutions Including the Clinical Center. Less ex- 
tensive quest lona I res administered in lORP 
revealed several dimensions which appear to be of 
practical and theoretical Interest for character- 
izing ward value orientation. 



256 



Obj ect Ives : 

The primary objective of this study is to charact- 
erize the value orientation prevailing in these 
psychiatric wards in terms of fourteen categories 
(called 'scales'). (The response items, which are 
in multiple choice format, are distributed among 
these mutually exclusive scales so that each subject 
is characterized by his set of scores.) An allied 
objective is the assessment of reliability over time 
of certain factors discovered in the early stages 
of the study. 

Work in FY '72: 

A test-scoring program has been written which edits 
and transforms the raw response items and computes 
a subject's scales. When data on enough subjects 
has been compiled and processed, certain statistical 
programs will be run. Factor analysis will be the 
principal analytical tool with correlation compu- 
tation and discriminant analysis playing lesser 
roles . 

Major Findings: 

Although a pilot analysis was conducted in 1969, a 
statement of major conclusions should be deferred 
until the newly acquired data has been analyzed. 

Proposed Course: 

See Work in FY '72. 

h'onors and Awards: None 

Publ i cat ions : None 



257 



Sprial No. 1 .hi 

1. Pata Management Pranch 

2. Mathematical anH 

Statistical J^ection 

3. Bethesda 

PHS-NIH 
individual Project Report 
July 1, 1971 through June 30, n72 

Project Title: Family Psychophys iol osy Project 

Previous Serial Number: None 

Project Leader: Dave VanSant/ nCPT 

Other Projess ional s : None 

Cooperating I'nits: Section on Personality f^evel opment. 

Adult Psychiatry Pranch, NIMH 

Man Years: 

Total : 0.20 
Profess Ional : 0.20 
Other: n.OO 

Project Description: 

Background: 

This project is concerned with anxiety in family 
interaction and its relation to the behavior of 
emotionally disturbed adolescent family members. 
Observable family behavior and therapists' infer- 
ences are supplemental with a new source of Infor- 
mation, the galvanic skin response of therapy 
session members. 



258 



Object I ves : 

The objective of this study Is to closely study the 
interactive behavior of families characterized by 
troubled adolescents. The recording of galvanic 
skin responses may be a valuable nev/ technique 
enabling therapists to intervene more successfully 
In family therapy sessions. 

Work in FY '72: 

We believe that the major programming effort in 
this study has been completed. Minor program modi- 
fications have recently been made in two of the 
four processing programs and data is continuing to 
be analyzed. Hue to the unusually difficult process 
Involved in data coding (done by the APR of NIMH), 
it is expected that the programs will continue to 
be run for quite some time to come. It also appears 
that minor adjustments in the programs will continue 
to be made as more results are generated and 
anal yzed. 

r'ajor Findings: 

As soon as enough family therapy sessions have been 
analyzed, a statement of major findings will be 
possible. Ife estimate that it will be about six to 
nine months before a report of major findings will 
be forthcoming. 

Proposed Course: 

See Work in FY '72. 

f^onors and Av/ards: None 

Publ icat Ions : None 



259 



Serial Mo . 7 . U "^ 

1. ^ata Management Pranch 

2. Mathematical and 

Statistical Section 

3. Pethesda 



PHS-N IH 
Individual Project Report 
July 1, iq71 through June 30, 1"7? 



Project Title: 
Previous Serial fJumber: 
Project Leader: 

Other Professionals: 

Cooperating Units: 

f^an Years : 

Total : n.iiO 

Profess ional : n . i»0 
Other: n.nn 

Project r>escr i pt ion : 

Objectives: 



Animal Oenetics 

None 

Fred Yamada^ ncPT 

None 

Laboratory Aids "ranch, PRS 



This study is concerned with the genetic analysis 
of characteristics which have application to bio- 
medical research. 



260 



Work Pone in FY '72: 

This project thus far has dealt with the analyses 
of reproductive patterns in genetically defined 
strains of mice, rats and guinea pigs. The major 
activity of this project has been an analysis in- 
volving 2'+ mouse, 12 rat and two guinea pig strains 
from data collected over a long period of time (11 
years). The analyses have included analysis of 
variance, multiple regression and Cal-Comp output 
for seven characteristics measured in each of the 
strains. Some preliminary analyses have been com- 
pleted on a project consisting of nine strains of 
mice studying the relationship betv/een age and 
reproductive patterns. Finally, a preliminary 
analysis has also been completed on a project in- 
volving the development of a standard mouse strain 
for drug evaluation. 

Major Findings: 

Not available at this time. This study is not 
yet complete. 

Proposed Course: 

To conclude the foregoing evaluations adequately. *S! 

Honors and Awards: None '■"'I 

Publications: None ii«<i 



261 



Serial Mo. 7.ttU 

1. Hata ^^anasement branch 

2. Mathematical and 

Statistical Section 

3. Pethesda 

PHS-MIH 
Individual Project Report 
July 1, 1P71 through June 30, 1P72 

Project Title: Interview Scheduling System 

Previous Serial Number: None 

Project Leader: Ray Hanner, f^CRT 

Other Professionals: Dave VanSant, DCRT 

Cooperating Units: Clinical and Professional Services 

Section, nn, CC 

Man Years: 

Total : n.n? 
Profess ional : n.07 
Others: n.OO 

Project Pescription: 

Objectives: 

The objective of the Automated Scheduling System 
is to design and produce an optimized schedule of 
interview periods for both the applicants (•'^or 
associate program) and the interviewers (Scientific 
staff at NIH). 



262 



VJork Done in FY '72: 

Major modifications and complete rewriting of seg- 
ments of the entire system v/ere necessary to meet 
the new specifications requested by the Clinical 
Center. Appropriate modifications were made to 
deal with the following specifications: 

1. Increase the number of interviev/ periods per 
day to 16 and a maximum of 32 interviews per 
appl i cant . 

2. Increase the number of applicants to 2000. 

3. Increase the one day applicants to 2 days 
should they have more than 16 interviews. 

Methods Employed: 

For a given applicant the interviewers requiring 
special appointments are scheduled. The remaining 
interviewers are then grouped by the building num- 
ber of the location where they will conduct their 
interviev^s. A building group is selected and the 
time schedules for the interviewers in the building 
are used to form a matrix. This is a matrix with a 
row for each of the 14 possible interviewers 
and a column for each of the seventeen possible 
time periods. The time periods are selected so ig] 

that they form a series of earliest available con- «« 
secutive periods. The scheduling scheme essentially '** 
schedules the least available interviewer, in the 
least available time period. it'* 

\lhen an interviev^er is scheduled in a particular ■'., 

time period the row is labelled with the column of i""" 

the time period, while the column is labelled with iSu 

the row number of the interviewer. J 



263 



The row routine is triggered if any of the row sums 
equal zero. It causes the algorithm to consider 
the next building group. If all the unscheduled 
building groups have been considered, it adds a 
column (time period) to the matrix and repeats the 
procedure. If the matrix is still unschedul eabl e, 
a new matrix is considered. This is achieved by 
considering the time period two periods later than 
the initial period as the new initial period. The 
procedure continues until all available time 
periods have been considered. If the matrix re- 
mains unscheduleabl e after all periods have been 
considered the scheduling constraints are removed 
and the interviewers are scheduled In any available 
time period. 

The column routine Is triggered by a zero column 
sum. Its operation Is similar to the rovj routine. 
The only difference is that it begins Its search 
by adding a column to the matrix rather than 
searching building groups. 

If the applicant cannot be completely scheduled on 
a given maxday (day with most appointments), the 
algorithm tries to schedule the applicant on 
another maxday. If no further day Is available, 
the day on which the greatest number of appoint- 
ments Is scheduled is selected. 

flajor Findings: 

'Major Finding' does not apply to this project, 
however, this system has run successfully before 
and It Is anticipated that this expanded version 
will run this year optimising the scheduling of 
Interv lew periods . 

Proposed Course: 

To adequately generalize the system for general use 

I'onors and Awards: None 

Publ I cat Ions : rJone 



264 



Serial No. 7.tt5 

1. Data Management Cranch 

2. Mathematical and 

Statistical Section 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Math/Stat Library 

Previous Serial Number: None 

Project Leaders: George Shakarjl, DCRT 

Bob Baird, DCRT 
Dave VanSant, DCRT 
Ui 1 1 iam Vincent, DCRT 

Other Professionals: None 

Cooperating Units: None 

Man Years: 

Total : 2.20 

Profess ional : 2 . 20 ,,1^^ 

Others: 0.0 dC!£ 

(1*1 

Project Description: ,«i«il 

Work in FY '72: |f* 

1. A nev; flathema t i cal and Statistical program -a; 

library manual was published. This hand book '*pi 
is designed to provide detailed documentation '"f' 
and run instructions on a collection of math- S 
ematical and statistical programs which are sJ 

most commonly used at fJ I H . This handbook has •""•"i 
been published in loose-leaf form in anticipa- 
tion of an ongoing updating program. Correc- 
tions, modifications, and additions to this 
handbook will be published at reasonable 
i n te rva 1 s . 



265 



In order to facilitate usage of the Math/?;tat 
programs, a large library of programs (excluding 
RMD's) v/as created on a resident disk pack. All 
programs on the pack are stored as fully re- 
solved load modules. This will eliminate link- 
edit time upon execution, thus there should be 
less cost per run. The library is a multi- 
member partioned data set which is periodically 
copied onto a back-up tape, so that if the data 
set is destroyed, it can be replaced easily and 
with minimum delay in user service. 

A nev7 version of the Piomedical Computer pro- 
grams (BMD) has been loaded on our private disk 
pack. f^pecial procedures have been developed 
and implemented to improve the execution time 
and turn around time by making it possible to 
run in A class rather than the P class or C 
class runs. The JCL needed for execution is 
also made simpler for the new version. 

Because of user demand, the Math/Stat Section 
requested and received the IBM-J;/3P0 Continuous 
System Modeling Program (CSMP) and loaded the 
object module on a private on line disk pack. 
CSMP is a problem-oriented program designed to 
facilitate the digital simulation of continuous 
processes on large-scale digital machines. The 
program has been tested and it provides an 
application-oriented language that allows these 
problems to be prepared directly and simply 
from either a block-diagram representation or a 
set of ordinary differential equations. 



266 



S ign I f i cance : 

The implementation and development of programs and 
procedures in the Math/Stat library would give NIH 
investigators easy access and simplified run pro- 
cedures to statistical and mathematical programs to 
be applied in the vast area of Piomedical Research. 

Proposed Course: 

Continue to develop and maintain a large program 
library that is easily accessible to users. ripvelop 
further procedures that would require even less 
programming knowledge by investigators to access 
I I brary programs . 

Honors and Awards: None 

Publ icat ions : None 






iHaiiil 



267 



Serial Mo. 7. UP 

1. Hata Mana^enent Pranch 

2. Mathematical and 

Statistical Section 

3. Rethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1<)7? 

Project Title: Nutrition Study of Laboratory 

Animal s 

Previous Serial number: None 

Project Leader: Fred Yamada, PCRT 

Dther Professionals: None 

Cooperating Units: Veterinary Resources Pranch, r»RS 

Man Years : 

Total: 0.30 
Profess ional : 0.30 
Others: 0.00 

Project Description: 

Objectives: 

A program has been established in the Veterinary 
Resources Pranch of the Pivision of Research Ser- 
vices to improve the nutritional status of labora- 
tory animals at NIH. In order to meet this 
objective a series of feeding trials have been 
conducted to determine the general nutritional 
requirements of various strains of laboratory 
rodents. Large amounts of detailed growth and 
reproduction data have been accumulated. 



268 



Work in FY '72: 

Most of the work completed thus far in this study 
has been in the area of generating and defining 
growth curves and least square analyses of the data. 
These analyses have provided a considerable amount 
of knowledge regarding the biology of the species 
invol ved. 

Major F ind ings : 

A more direct result of these studies has been the 
development of an improved open formula rodent 
ration that is available to NIH investigators. 

Proposed Course: 

Analysis has begun and will continue to be applied 
to large amounts of pathological, reproduction and 
growth data collected from canine colonies at the 
Animal Center at Peltsville. The objective of this 
evaluation is to improve the general status of the 
NIH canine blood donor and production colonies. 

Honors and Awards: None 

Publ icat ions : None 



269 



Serial No. 7.tt7 

1. Data Management "ranch 

2. Mathematical and 

Statistical Section 

3. Rethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 50, 1P72 

Project Title: Chemistry Pata Handling and 

Fvaluation SystP'm 

'Previous Serial Number: None 

Project Leader: George Shakarji, ^CRT 

Other Professionals: None 

Cooperating Unit: Endocrinology Branch, NHL I 

Man Years: 



Sotal : 


0. 


.kO 


Profess lonal 


: n. 


.kO 


Others : 


n. 


,00 



Project Pescription 
Object i ves : 



The initial objective of this system is to pro- 
vide a massive storage and retrieval capability 
for the chemistry data of the Endocrinology 
Pranch, NHL I . The data is stored on tape exe- 
cuting programs that would produce variable 
length records of demographic information and 
test results which are linked to form patient 
files. 

These programs are designed to maintain the 
chemistry files by a) adding new demographic 
and test records to the patient file; b) 
making appropriate changes and deletions; c) 
deleting spurious Information. 



2 70 



2. Another objective of this study Is to statis- 
tically evaluate the results of the laboratory 
analyses of the given samples from an already 
categorized and easily accessible chemistry 
file. 

Methods Employed: 

1. The patient number. In addition with the first 
and second diagnosis codes form the primary 
keys that link demographic and test records 
forming the patient's file. Accession codes 
and the test date form secondary keys that link 
the separate laboratory tests performed on the 
same sample of body fluid. Each demographic 
record contains an updatable counter indicating 
the number of test records stored for that 
particular patient. 

2. Methods to be employed for statistical evalua- 
tions will be explored after the completion of 
the chemistry data on the defined group of 
pat ients. 

S ign I f Icance : 

The patient chemistry file provides a data base for 'JJS 

easy access of patient chemistry information In gS 

addition to providing a data base for scientific i«*| 

evaluations in biomedical research areas. The pro- ""■ 

cedures for building each patient file were designed .mn 

in a form suitable for easy access of chemistry data |,^.,„ 

and for statistical analyses. p^, 

1 41 ■ 

Proposed Course: .nrf' 

40" 

Study and design a system that would provide re- """I 

curring output for review by the scientific staff "■" 
of MHLI and study statistical approaches with '"""" 

statisticians for scientific evaluations of the 
data. 

Honors and Awards: None 

Publications: None 



271 



Serial ;Jo. 7.U8 

1. Data Management Branch 

2 . Appl led Systems 

Programming Section 

3. Rethesda 



PHS-nill 
individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Dysllpldemia Study 

Previous Serial .Number: fione 

Project Leader: Penny Progan, PCRT 

Other Professionals: Roger Dai ley, DCRT 

Cooperating Units: Lipid fletabolism Branch 

flan Years: 



Total : 


0.8 


Profess ional : 


n.8 


Others : 


0.0 



Project Pescription: 

Objectives: 

The purpose of this project Is to design and implement 
a system for maintaining a computer file v;ith basic 
diagnostic information on U\\] patients vjho have 
dys 1 i p i demi a, their relatives and some control 
patients. The file Is composed of identifying infor- 
mation on each patient and the results of lab tests 
and physical examinations by physicians. Various 
selective listings and statistics will be produced 
from the data. 

Methods Fmployed: 

Data will be collected and corrected on punch cards, 
but stored on 9 track tape. COBOL programs have 
been written to update the file and produce desired 
repor ts . 



272 



Sign i f Icance : 

The system will provide a more rapid means of keeping 
the progress of each patient up to date. This will 
yield more current statistical data as well as quick 
reference lab reports for each examining physician. 

Proposed Course: 

The system has been defined and programmed, except 
for the processing of some family history data. The 
ability to maintain the latter data will be pro- 
grammed. JCL to process each phase of data handling 
is now being set up and documentation is being 
completed. 

Honors and Av;ards: None 

Publ icat ions : None 



273 



Serial :io. 7.U9 

1. Data Management Pranch 

2 . Appl led Systems 

Programming Section 

3. Cethesda 

PHS-;JIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: M.IIyoiner Arthritis Study 

Previous Serial IJo.: ilone ' 

Project Leader: Roger Dai ley, DCRT 

Other Professionals: Anne nallagher, DCRT 

I'lary Lee Dante, DCRT 
George Dobenecker, DCRT 
Heorge Roberts, DCRT 

Cooperating Units: ;^JIAf1D-ARB 

flan Years : 

Total : 0.20 

Profess iona 1 : 0.18 
Others: 0.02 

Project Description: 

Obj ect i ves : 

The purpose of this project is to design a system for 
editing, updating and reporting on clinical observa- 
tion, laboratory data and microbial studies files. 
Data for these three files will be supplied by the 
cooperat i ng un i t . 

f'ethods Employed: 

Because of the small size of the files and the small 
number of errors found, the files viere stored on an 
on-line data set and updating v;as done manually. The 
laboratory data and microbial studies data files were 
combined into one file for the ease of reporting. 



274 



Major Findings: 

None . 
S Ign i f i cance : 

None . 

Proposed Course: 

Data files and computer programs turned over to 
cooperating unit. 

Honors and Awards: None 
Pub! icat ions : None 






mm'' 



2 75 



Serial No. l.'^O 

1. Hata Management "ranch 

2 . Appl ! ed 5?ystems 

Programming 5^ectlon 
3. Pethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1072 

Project Title: MIH Library Operations 

Previous Serial No.: None 

Project Leader: Roger Pailey, PCRT 

Cooperating Unit: NIH Library 

Man Years: 

Total: : 0.20 
Profess ional : 0. 20 
nther: 0.00 

Project Pescription: 

Object i ves : 

After the NIH Library's existing card system was con- 
verted to an online VYLPUR system, creating a large 
'/YLBUR data set, it became an immense task to maintain 
via the "YLBUR terminal. The purpose of this project 
is to write a program or programs which would do the 
maintenance and retrieval of information from this 
','YLBUR data set. 



276 



Method Employed: 

A COBOL program was written to delete all records on 
the NIH Library's master '/YLBDR file, which met the 
criteria supplied by a parameter card and predefined 
condition within a key record on the file. All 
reporting was handled by IRS querier. 

Proposed Course: 

To provide program maintenance and other programming 
as requested. 

Honors and Awards: None 

Publications: None 



277 



Serial No. 7.'^! 

1. Data Management Pranch 

2 . Appl led Systems 

Programming Section 

3. Pethesda 

PHS-MIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Collection and Analysis of 'Genetic 

Hata 

Previous Serial Mumber: None 

Project Leader: George f^obenecker port 

Other Professionals: None 

Cooperating Units: N|A|D/LVn 

flan Years: 

Total: n.30 
Professional : n. 25 
Other: n.05 

Project Hescription: 

Object i ves : 

The major objective of the project was to provide r>r. 
Powe v/ith a computer system that would collect 
genetic data and mal<e the information available for 
on-1 ine anal ys i s . 

Methods Employed: 

Forms were designed so that the laboratory technicians 
could record the desired information. l-fhen a suffi- 
cient number of forms have been coded they are sent 
to the Key Punch Section of PORT to be loaded onto a 
WYLBUR data set. (This is done to facilitate the 
correction of errors generated by the edit/update 
step.) A technician logs onto ''YLPI'R and executes 
the edit/update programs. When the run is completed 
any errors are corrected on the input './YLRl'R data set 
and the edit update step is re-executed. As a by 
product of the update a subset of the master file is 
written onto an on-line disl< pack that is available 
to CPS. 



278 



Major Fi nd ings : 

Making the data available to the researcher for 
analysis via CPS gjreatly enhances his ability to use 
the data without further programmer intervention. 

Significance: 

None . 

Proposed Course: 

Support any additional requests by Dr. Powe to improve 
or extend the system in areas that he deems necessary. 

Honors and Av/ards : None 

Publ i cat ions : None 



IBx' 
"■If 



inml' 



2 79 



Serial No. 7. ■^2 

1. Data Management Branch 

2. Applied Systems 

ProgrammlnK Section 

3. Pethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Transposing '-'YLBUR Data into a Format 

Compatible with Publication Requirements 
of the Information (Including Heading, 
Footings, Page Numbers, etc.) 

Previous Serial flumber: None 

Project Leader: Heorge '^obenecker, PCRT 

Other Professionals: None 

Cooperating Units: DHEW Office of Education 

'^an Years: 

Total: 0.25 
Profess ional : 0.20 
Other: 0.05 

Project Description: 

Objectives : 

The objective of the project was to give the Office 
of Education the ability to collect the text of a 
manual on IIYLBUR and then execute a program that 
would output the text in a format compatible with 
the publication requirements of the Office of 
Education. 



280 



Methods Employed: 

The text is collected, updated and maintained on 
WYLBUR. '/hen the user wishes to publish the infor- 
mation it is listed via the !'/YLBIIR command 'list 
offline'. The listing is reviewed and the line number 
of the last desired line on each page is collected on 
a V/YLBUR data set. The text format program is then 
executed. it supplies the heading and footing infor- 
mation, page number, continuation information etc. 
The number of lines printed on a page is controlled 
by the line numbers read from the second '■'YLRI'P Hata 
set. 

Major Findings: 

flone . 

S ign i f i cance : 

A slight modification of the format program written 
for the Office of Education was used to format the 
text for the publication of the PCRT/nMB Math/Stat 
documentation text. 

Proposed Course: 

Support additional requests to format text for pub- IjJJ 
lication when that text is normally collected and *]j 
updated using WYLBUR. '*" 

iiiiiii 
Honors and Awards: None 

Publications: None ,aa 



wISl 



281 



Serial Mo. 7.^3 

1. Pata Manasenent branch 

2 . Appl led Sys tens 

Programming; Section 

3. ^ethesda 



Project Title 



PHS-MIH 
Individual Project Report 
July 1, in71 through June 30, 1^72 

Administrative Information 
Retrieval System 



'^rev'ious Serial '.'unber: 
Proj ec t Leader : 

'^ther Investigators: 
Coope rating I'n i ts : 
"an 'r'ears : 



None 

Vivian Pelhan, r^CP.T 

'lone 

Administrative i^^fice, ncPT 



Total : 


1, 


,0 


"rof ess i onal 


: f^ , 


_ n 


^ t h e r : 


0, 


!o 



'roject description 



O^j ec t i ves : 



The 
for 
data 
sat i 
and 
pers 
add i 
mass 
f iel 
gene 
c 1 ud 
tota 



purp 
s tor 
. T 
ona 1 
ret r 
onne 
t i on 
cha 
ds o 
rat i 
ing 
Is, 



ose of this project is to design a system 
age and retrieval of personnel and budget 
he system is to be written as a conver- 

system for quick and easy data maintenance 
ieval of information by administrative 
1. The system is to allow the follovnng: 
, modification and deletion of information; 
nges of variable data (i.e., change salary 
f all records to reflect a pay raise); 
on of recurring and one-time requests in- 
summaries of specified data fields (i.e., 
averages, counts, etc.). 



2S2 



Methods Employed: 

Three separate systems (Personnel, Training and 
Space) are to be developed, one at a time, using 
as many of the programs and subroutines written 
for foregoing systems as possible. RMAG programs 
are used to generate a COBOL update program and IRS 
input for one-time queries. Each system will con- 
sist of a CPS main program to prompt the user with 
various options and call the corresponding pro- 
cedures to: generate transactions for updating the 
master file; update and list the salary table and 
generate a transaction to change all salaries 
accordingly; update and list the organization table 
and generate a transaction to reflect these changes 
on the master file; change the fiscal year in all 
records; list the master file; query the master 
file using stored IRS input for recurring reports 
or prompt the user for selection and sort criteria 
and report specifications. 

All data fields are edited, and salary and organi- 
zation information is verified against the respec- 
tive files before data is allowed to be written on 
the transaction file. All procedures called by the 
main program consist of a series of prompts, guiding 
the user in operation. 

Once a user has completed an option all programs 
which have to be submitted to the batch stream are 
automatically scheduled for him. 

Sign i f Icance : 

The system allows non-computer personnel to main- 
tain files and retrieve information from these with 
little effort. Many hours were saved by generating 
programs using RMAG. 

Proposed Course: 

The Personnel system is operational and is to be 
evaluated for future modifications. Several insti- 
tutes have requested that such a system be developed 
for them. 

Honors and Av/ards: None 

Publ icat ions : Mone 



283 



Serial No. 7.5tt 

1. Data Management Branch 

2. Appl led Systems 

Programming Section 
3. Bethesda 

PHS-;JIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: NMR (Nuclear f'agnetic Resonance) 

Literature Retrieval System 

Previous Serial .Jumber: fNione 

Project Leader: Vivian Pelham, DCRT 

Cooperating Units: Section on flolecular Biophysics 

Laboratory of Physical Biology, 
U I AMD 
Preston Technical Abstracts Co. 

flan Years : 

Total: n.i5 
Profess ional : 0.13 
Other: 0.02 

Project Description: 

Obj ect i ves : 

The purpose of this project is to maintain a data 
base of abstracts of literature on .UlR and provide 
retrieval of these by year, author, journal, subject 
classification code and combinations of these. 



284 



Methods Employed: 

The system consists of two COBOL programs, an up- 
date and a retrieval program. Abstracts of NMR 
literature published by the Preston Technical 
Abstracts Company are scanned for key v/ords and 
authors (if more than U) by lilAMD personnel and 
the significant information is punched onto cards 
for input to the update program. Searches are run 
against the file by DCRT as requested. 

Sign i f icance : 

Several agencies receive a copy of the updated 
master file and have the search program running on 
their own computers. 

Proposed Course: 

Several revisions are to be made to the system. 
The subject classification codes have been redefined 
to allow unique codes for each nucleus. This will 
entail recoding of present codes and expanding many 
which now encompass several nuclei. Journal infor- 
mation will be expanded to include volume and page 
numbers. Only three authors are to be recorded and 
the key words v/i 1 1 be expanded to allow a more „a, 

descriptive summary. Cm 

■>« 

"./hen these revisions have been completed a fully •««' 
documented system vn 1 1 be turned over to and 
operated by fJIAMD. 

i:!t:iJ 

Honors and Awards : None i i ' 

nil'"" 

Publications: None T: 



«SII 



285 



Serial fJo. y.^^S 

1. Pata Management Branch 

2. Applied Systems 

Programming Section 

3. Pethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1^72 

Project Title: NHL I Contract Grant System 

Previous Serial Number: None 

Project Leader: George Oobenecker, PCRT 

Other Professionals: Roger Halley, PCRT 

Cooperating Units: National Heart and Lung Institute 

Man Years: 

Total: 0.25 

Professional : 0.25 
Other: 0.00 

Project Pescription: 

Object i ves : 

The system was designed to collect, edit, update 
and report on data pertaining to the contract grants 
awarded by NHL I . 

Methods Employed: 

Tv/o files are maintained by the system, the con- 
tract grant master file and the detail financial 
file. Forms were designed to support the data 
collection of the information necessary to maintain 
the files. The data is read into the system and 
the cards used to update the master file are sepa- 
rated from those used to update the financial file. 
The master file is updated via the standard update 
program which appends the information to the proper 
master file record and then passes the updated master 
file to an edit program. After the complete master 
file has been updated the financial data is edited 
and the valid financial cards are used to update 
the financial master file. In the final step the 
newly updated financial file is passed and the 



grants which have had changes to their award amounts 
have those changes appended to the Fiscal Year 
totals carried in the master file records. The 
changes are printed on a report used by NHL I to 
verify that the correct changes were made. 

Major Findings: 

None 
S ign i f i cance : 

None 

Proposed Course: 

Support the conversion of NHLI's other grant 
systems, which are currently processed by EAM 
equipment/ to computer systems. 

Honors and Awards: None 

Publications: None 






287 



Serial r'o. 7.56 

1. Data rianap;empnt '-ranch 

2. Documentation and 

Supper t Sect i on 

3. nethesda 



'roj ect Title: 



PHS-V'IH 
Individual Project Report 
July 1, 1971 throu,s:h June 30, 1972 

Tape Conversions 

...ind P 1 ate i lak i nrr 



Previous Serial r'l.'mber 
Project Loader: 
Other Professional: 
C'-'Onera t i n-^ l.n its: 
■ ; an Y .-^ a r s : 



None 

G '.^ o r ;?; e S c h u 1 z e 

None 

flono 



Total : 


0.3 


'"'rof ess i on:H 


: 0.3 


Others: 


COO 



Droject ," nscr i pt i on : 

Obj ect i ves : 

To provide a capa''ility o^ t rans 1 at i n,-^ iriultifil'^ 
tapes. Tabled and non-labled, 7 and 9-track, from 
E5C01C to ASCII-S, from ASCI I -S to FSCDIC, and fro" 
binary to 360 compatible binary. Each o-^ these was 
a spec i f i c case. 

Methods Employed: 

f-inr basic assembly lan-ruap-e rro.sTam was written and 
then modified to acconmodate i^ach specific reou'^st. 

' '■■?/] or F i nd i n;-:s : 

Even though 05 release 20 can handle ASCll-8 to 
ESCiniC and E'CIDIC to ASCll-8 translation, a special 
pro?;ram must be used if record leneth <^reater tiian 
2000 bytes is used. 



288 



Signi f icance: 

A greater flexibility of processing data gathered 
through the smaller computers on the NIH campus. 



Proposed Course: None 
Honors and Awards: None 
Publ icat ions : None 






mm 



289 



Serial No. 7.57 

1. Data Management Branch 

2. Documentation P: Systems 

3. Support Section 
k . Be tbesrla 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Nursing Activities Study-13 East 

!>' Retrieval System 

Previous Serial Number: None ~" . — - 

Project Leader: Dennis George, DCRT 

Other investigators: Georg J. Schulze, DCRT 

flan Years: 

Total: 0.1*5 
Profess i ona 1 : 0.35 
Others: 0.10 

Project Description: 

Objectives: 

To produce statistical listings of nursing activities 
from coded forms. This was a one-time-only pro- 
gramming effort. 

Ilethods Employed: 

Standard system design and programming techniaues 
have been employed. 

f'ajor Findings: None 

Proposed Course: None 

Honors and Av;ards: None 

Pub 1 i ca t i ons : None 



290 



Serial No. 7.58 

1. Data Management Branch 

2. Documentation and 

Support Section 

3. Bethesda 



PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Opportunity Skills System 

Previous Serial Number: None 



Project Leader: 
Other Investigators 
Cooperating Units: 
Man Years: 



Georg J. Schulze 

None 

PM 



Total : 


0. 


,uo 


Professional : 


0. 


,30 


Other: 


0. 


,10 



Project Description: 

Ot>ject i ves : 

Implementation of a data retrieval system givin.p the 
office of personnel management the capability to 
interactively search and update a skill file. To 
determine the best format for this file and to con- 
vert the present documents into computer readable 
format and to create this skills file. 

Methods Employed: 

Standard information system design and CPS ^ PL/1 
programming techniques have been used. 

Major Findings: 

None 



mti 



ii«i>' 



291 



Proposed Course: 

0PM personnel will assume responsibility for file 
maintenance and the runninr or the System efte.r 
development. I'CRT will provide limited sunr-ort. 

Monors and Awards: None 

Pub 1 i cat i ons : None 



292 



Serial No. 7. 59 

1. Data Management Branch 

2. Scientific Applications 

Sect ion 

3. Bethesda 



PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Printing Pressmen Study 



Previous Serial Number: None 

Project Loader: Anne Gallagher, DCRT 

Other Professionals: Leo Burkett 

Cooperating Units: llliness and Injuries Surveillance 

Branch, NIOSIl" 

Man Years: 

Total: 0.17 
Profess i ona 1 : 0.17 
Others: 0.00 

Project Description: 

Obj ec t i ves : 

Preliminary studies have shown unusual mortality 
among pressmen for certain diseases (ex. emphysema, 
cancer of the buccal cavity and pharynx and es- 
ophageal cancer). The purpose of the present 
study Is to help clarify the specific disease 
problems of printing pressmen by looking at pro- 
portionate mortality. 

Methods Employed: 

The count program read in a partially recoded file 
and created three sets of counts. The first set 
of counts was broken dov/n by Br,(^, speciality, cause 
of death and region. The second set was broken 
down by age, speciality, locale, special cause of 
death and nativity (the special cause of death being 
a combination of contributing and underlying causes 
of death). The third set of counts was broken dov;n 
by age, cause of death and activity. 



293 



The calculation program produced two types of tables. 
The first table type compared the U.S. white male 
mortality counts (using same age, cause of death 
and region categories) with the first set of counts 
and calculated observed and expected deaths, 
variance, ch i square and significance values. The 
second table type compared the third set of counts 
with the second set (specific spec i a 1 i t y- 1 oca 1 
pressmen with all pressmen) and calculated the same 
f i ve val ues . 

The output program, through the use of parameters at 
the front of each set of values, printed the five 
values per line in table form with the appropriate 
head I ngs . 

Major Findings: 

None to date since analysis of output has not begun. 

Significance: None 

Proposed Course: 

These programs are to be generalized. In particular 
the calculation and output programs to a 1 1 ov; for 
their utilization In similar studies. 

Honors and Awards: None 

Publ I cat Ions : None 



294 



Serial No. 7.60 

1. Data Management Branch 

2. Documentation and 

Support Section 

3. Bethesda 



PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 



Project Title: Clinical Laboratory Beta File 

Convers i on 

Previous Serial Number: None 

Project Leader: Joseph L. Campbell 

Other Professionals: Dennis George 

Cooperating Units: None 

Man Years: 

Total: 0.35 
Profess ional : 0.35 
Others: 0.00 

Project Description: 

Object i ves : 

The purpose of this conversion was to provide the 
Clinical Center research investigator v/i th a quickly 
retrievable data base of patient test results. Both 
chemistry and hematology test results are recorded 
as they occur on the clinical laboratory computer 
system and these results are made available to the 
Clinical Center investigator on a daily basis. For 
analysis, however, the structure and volume of the 
file needed to be handled differently. A new method 
would provide easily retrievable data and enable the 
investigator to analyze effects of various clinical 
trials as they relate to his patients test results. 



295 



Methods Employed: 

Since the collection and daily reporting of labora- 
tory data was acceptable, the Data Management Branch 
in collaboration with the clinical laboratory concen- 
trate on converting the bit encoded laboratory file 
to a character oriented data base for quick retrieval, 
A system for further fragmenting this data base for 
individual research investigators was then developed. 
This provides each investigator with test results of 
those patients of interest to him and eliminates the 
need for him to search the entire clinical laboratory 
f i le. 

Major F i nd i ngs : 

The use of a structured character oriented data base 
provides the clinical investigator the easiest method 
for analyzing patient test results. 



Honors and Awards 



None 



Publ icat ions: 



None 



296 



Serial Mo. 7.61 

1. Data Management Branch 

2. Documentation and 

Support Section 

3. Bethesda 



PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 



Project Title: Commercial Timesharing Users Survey 

Previous Serial Number: None 

Project Leader: John A. Haggerty, Jr. 

Other Professionals: J. Gilliss 

T.J. f^eavey 

Cooperating Units: All B/i/D's with Contracts 
fian Years: 



Total : 


0.60 


Profess ional 


: 0.30 


Others: 


0.30 



Project Description: 
Obj ect i ves : 



1. Determine the needs of time sharing users - 
how, why and for what specific reason(s) or 
appl i cat ion( s ) are they using these services. 

2. Apprise them, that at renewal time justification 
for continued use of such services vn 1 1 be re- 
quired by GSA as they will only renew those 
contracts which provide service(s) that cannot 
be matched either in-house or at any other 
Government installation (including CSA's own 
service in Atlanta, Ga.). 

3. Provide information on what is available In PCRT 
including Hardware a Software, Terminals, etc. 
and encourage use of DCRT facilities where 

poss i bl e. 



297 



k. Rc-duce tho number of cotnr.ierc i a 1 timesharing; 
contracts in effect and try to p:et users of 
multiple services to reduce their use to the 
one or two most necessary services. 

Method Employed: 

Interviewed the majority of users to rather data as 
per objectives #1, 2 and 3 Set up a form to record 
information supplied in the interviews and also to 
record the types of terminals and other peripheral 
equipment installed at the user's location. 

f1a jor F i nd i ns;s : 

The majority 0"f^ users required an interactive system 
which was not available in PCRT. Some of these users 
had either never used '~>CPT or not since the presf^nt 
system wos i ns tal 1 ed,and thus did not know w^at was 
available from PCRT. Other factors which users cite' 
as important reasons for continuinr such s'^rvices 
included: ease of prorrainmi n.r^ in Pasic Lanruares; 
availability of pro"-ram libraries; j^raphs, charts 
and nap making capabilities; larp;e core rf nu I rcments ; 
shared data bases collected over lont^ periods of 
time; and time and cost restrictions Involved in 
convertinp to the DCRT computer facility. 

3 i gn I f i ca nee : 

Since the survey was conducted the number of con- 
tracts has been reduced from 27 to the presf^nt 10 
which now includes OSA In Atlanta. After learninr 
what \:as available in DCRT many users are no\'' 
doing more of their work In-house and the nur^ber of 
users of commercial timesharinr services has been 
reduced from a high of 2h'i to tho present 105. 

Proposed Course : 

To follow-up at two or three month intervals by 
checking time sharing bills. A questionnaire was 
designed and mailed to all conimercial time s^^arin^r 
users requesting information to justify their 
continued use of such services/alonp" with auestions 
asking each user to estimate, i ■'^ possible, tho cost 
and time involved to convert to ncPT facilities. 

fionors and Awards: None 

Pub 1 I cat ions : None 



298 



Serial No. 7.62 

1. Data Management Branch 

2. Scientific Applications 

Sect i on 

3. Brthesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Computer Analysis of Radioligand 

Assay and Radioimmunoassay data 

Previous Serial Number: None 

Project Leader: Lee McBride, DCRT 

Other Professionals: Peter Vogel, DCRT 

Cooperating Units: David Rodbard, M.D., N I CHD 

Reproduction Research Branch, NICHD 

Man Years: 

Total 0.29 

Professional: 0.2 9 it««)|i 

Others: 0.00 fr" 

lUJllll 

t" 

Project Description: |«'"" 

Ob ject i ves : 

Researchers at NIH who perform tests requiring ;;|,||,; 

analysis of radioimmunoassay data used a system of J[, 

fifty programs on non-NIH timesharing computers. f;;, 

DMB has developed FORTRAN programs that combine nf' 

many of the previous programs for relatively in- '"",] 

stantaneous and inexpensive results. '* 

Methods Employed: 

Analysis of radioimmunoassay and competitive protein 
binding assay data. This method uses logit and log 
transformation to obtain a linear dose response curve 
followed by unweighted and iterated weighted least 
squares regression analysis (similar to probit 
analysis). The standard curve is plotted by computer, 
and potency estimates and confidence limits are 
obtained for unknowns. 



299 



Major Findings: 

Cut computer utilization costs by 80 percent. 

S i gn i f i cance : 

Radioimmunoassay programs are cheaper to run and are 
more universal when converted from BASIC to FORTRAN. 
This enables researchers to use the system (with 
minor modifications) at most computer centers. 

Proposed Course: 

Continue improving main program and convert more 
BASIC RIA programs to FORTRAN. 

Honors and Awards: None 

Publ i cat ions : 

The system of programs has been made available to 
the scientific community thru the National Technical 
Information Office. (NTIS) 



300 



Serial No. 7.65 

1. Data Management Branch 

2. Scientific Applications 

Sect i on 

3. Bethesda 



PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 



Project Title: 
Previous Serial Number: 
Project Leader: 
Other Professionals: 
Cooperating Units: 

Man Years: 

Total: 0.60 
Profess i onal : 0.60 
Others: 0.00 

Project Description: 

Object I ves : 



Cutting Oil Mi St Study 

None 

Frances Bauer, DCRT 

Ray Leckrone, 1 ISB, NIOSH 

Illness and Injuries Surveillance 
Branch, NIOSH 



1. The purpose of this study is to determine whether 
men exposed to cutting oil mists during metal 
machining operations experience unusual disease 
patterns In comparison with a well-defined non- 
exposed group from the same Industry. 

Methods Employed: 

The original card image personal and work history 
data consisting of from 3-15 cards per subject, 
have been thoroughly edited and reformatted Into 
one 700 byte record per worker. Where gaps indicat- 
ing a missing ticket in the employment record were 
detected, and where dates were missing, estimates 
were made to complete the record. Finally, vital 
statistics from death certificates were added to the 
file. This file will be used as Input for the pro- 
grams to produce Frequency Tables and Life Tables. 



301 



Major Findings: 

The system of editing programs has eliminated many 
transcription and consistency errors from the 
original records, and occasionally entire records 
were eliminated as in the case of female workers. 

S i gn i f i cance : 

By eliminating errors in the data, this preliminary 
work ensures that the statistics resulting from 
future analytical programs will be accurate. 

Proposed Course: 

During the coming year attention will he given to 
analyzing the cause-specific mortality patterns 
of workers exposed to several levels of cutting 
oil mists and the results compared with those for 
the non-exposed group. At a later date. Information 
on disease histories, physical examinations, smoking 
habits, and pulmonary function tests which have been 
compiled on certain subsets of this population will 
be analyzed. 



Honors and Awards 



None 



Publ i cat ions 



None 



302 



Serial No. 7 . 6U 

1. Data Management Branch 

2. Scientific Applications 

Sect ion 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Surgery Branch Edit and Update 

System 

Previous Serial Number: None 

Project Leader: Catherine Staneck, DCRT 

Other Professionals: Myrtle Morris, DCRT 

Cooperating Units: Surgery Branch, NCI 

Man Years: 

Total: 0.11 
Professional : 0.02 
Other: 0.09 

Project Description: 

Obj ect i ves : 

This project will attempt to improve and consolidate 
the computerized data collection procedures of the 
Surgery Branch in support of their research projects. 

Methods Employed: 

A terminal entry program and a file structuring and 
edit program are being generated semi -automat i ca 1 1 y 
for each of nine data types to be included in a 
single data base. The data entry and file building 
will be linked together to allow a Surgery Branch 
technician with a minimum of programming skills to 
maintain the data base. 

Major Findings: 

An evaluation of the project will not be accomplished 
for several months. 



303 



S i gn i f i cance : 

This project Is in the early stages of planning and 
its significance cannot be assessed at this time. 

Proposed Course: 

Continued programming generation to support this 
proj ect . 

Honors and Awards: None 

Pub 1 i ca t i ons : None 



304 



Serial No. 7.65 

1. Data Management Branch 

2. Scientific Applications 

Sect i on 

3. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 through June 30, 1972 

Project Title: Aortic Valvular Disease Study 

Previous Serial Number: None 

Project Leader: Catherine Staneck, DCRT 

Other Professionals: None 

Cooperating Units: Cardiology Branch, NHLI 

Man Years: 

Total: 0.16 
Professional: 0.16 
Other: 0.00 

Project Description: 

Obj ect i ves : 

This study will attempt to define prospectively 
pre-operat i ve parameters which seem to indicate in- 
creased operative risi< and those which indicate that 
irreversible myocardial dysfunction has occurred. 

Methods Employed: 

Data is being collected on patients prior to cardiac 
surgery and six months after surgery. These assess- 
ments consist of clinical, exercise and catheriza- 
tion laboratory studies. Pre- and post-operative 
data will be compared to determine operative success 
or failure. Programming has been completed to 
facilitate data collection and storage using general 
purpose editing routines available in the Data 
Management Branch. This procedure is designed to 
automatically connect data collection and file 
building in such a way as to enable a secretary in 
the Heart Institute with regular typing skills and 
a minimum of terminal instruction to manage this 
system. 



305 



Maj or F i nd i ngs : 

This project is in its earliest stages of data 
collection. The first six-month follow-up data 
vjill not be available until July and analysis will 
beg in at that t i me. 

S i gn i f i cance : 

Despite a varied data format including approximately 
one hundred forty different test results to be re- 
corded for each patient, a file structuring and 
specific edit program has been generated for this 
data vnth a minimum of programming effort using a 
general purpose edit generator. 

Proposed Course: 

A second phase of programming support of this project 
will include pre- and post-operative data comparisons 
and a correlation between pre-operat i ve factors and 
post -opera t i ve success or failure. 



Honors and Awards 



None 



Publ i cat ions 



None 



306 



Serial Ho. 7.66 

1. Data Manasemont Branch 

2. Scientific Applications 

Section 
5. Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1971 throu.e;h June 30, 1972 

Project Title: Low Birth V/eight for Ape Study 

Previous Serial Number: None 

Project Leader: Anne Gallagher, DCRT 

Other Professionals: None 

Cooperating Units: Eoidemiology <'i Biometry 

Branch, fJICHD 

Man Years: 

Total: 0.25 

Professional : 0. 25 , 

Other: 0.00 ! 

I 

I 

Project Description: , 

Objectives: ' 

To define a group of women prone to delivering a 

baby vn th low birth weight for age. Factors being 

tested are gestation period birth weight of baby, i 

combined with number of previous children born to | 

the mother, age of the mother, date of conception 

and month o^ delivery. 

Methods Employed: 

The original file contained records for half of all 
live births in 1958. Multiple births were eliminated 
as were records where the state of residence or the 
state of occurrence was one of Ik states not having the 
date of last menstrual period (needed to determine gesta- 
tion period) creating a file of 1,188,851 records. 
The data was recoded and a cobol program used to 
divide the data by sex. Five major table types 
were to be developed vn th each table having a 
different set of variables. Counting the data into 
appropriate categories was seen to reouire more core 



307 



than is available hence the data was first sorted on 
major categories to decrease the number of dimensions 
in the counting arrays. Each count program produced 
arrays containing counts, sums of birth weights, sums 
of birth weight quantities squared, sums of gestation 
period values and sums of squared gestation period 
values. Total table arrays were also accumulated in 
the count programs. A second set of programs pro- 
duced arrays of counts, proportions, cumulative pro- 
portions, means and standard deviations as squared 
for each table. It also created arrays for subtotal 
tables. The output program required a parameter 
card indicating table number, sex, etc. It is able 
to print output for each of the five table types. 

Major findings: 

None to date since analysis of data has not begun. 

Significance: None 

Proposed Course: 

Similar tabulations will be done in the future with 
changes In categories as indicated by output from 
this phase of the study. 

Honors and Awards: None 

Publ i cat ions : None 



308 



PUBLIC HEALTH SERVICE - NATIONAL INSTITUTES OF HEALTH 
DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 
Report of Program Activities 
July 1, 1972 through June 30, 1973. t<3j||i 






I III IN 
In* III 



ANNUAL REPORT 
TABLE OF CONTENTS 

Page 

DIRECTOR' S SUMMARY 1 

Summary of the Associate Director 5 

Summary of the Assistant Director 7 

Research Project Reports 

1.1 Visual and Biological Shape 9 

1.2 Clinical Center Admissions 11 

1.3 Computer Acquisition of Physician Orders ■ 13 

1.4 Automated Processing of Medical Language 15 

1.5 An Information Processing System for Pathology Data 17 

COMPUTER CENTER BRANCH 

Summary 19 

Research Project Reports 

2.1 Computer Representation and Manipulation of Macro 

Molecular Electron Density Surfaces 25 

2.2 Comprehensive Communications Control 28 

2.3 Remote Computer Graphics 31 

2.4 PDP-10 Accounting System 33 

2.5 Implementation of the Cambridge Crystal File 35 

2.6 PDP-10 Systems ■ 37 

2.7 Macro Molecular Display and Manipulation 40 

2.8 On-line File Processing System ■ 43 

2.9 Virtual Machine Facility 45 



Page 

2.10 DATASTOR Program 47 

2.11 HASP/Shared Spool 49 

2.12 360 Systems Development 52 

2.13 WYLBUR Text-editing System 54 

2.14 CPS - Conversational Programming System 57 

2.15 Optical Document Processing System 60 

2.16 Computer Output Microfilm 62 

2.17 Installation of Time Sharing Option (TSO) of OS/360 MVT 64 

2.18 INTERFACE (technical communications) 66 

2.19 Computer Training Program 68 

2.20 User Support and Communications 70 

2.21 Computer Center Users Guide 72 

LABORATORY OF APPLIED STUDIES 

Summary 75 

Research Project Reports 

3.1 Biomathematics and Statistics 79 

3.2 Evaluation of Computer-assisted Systems for Patient Care 82 

3.3 Computer Methods in Cardiologic Research 86 

3.4 Statistical Research in Clinical Pathology 88 

3.5 Applied Mathematics Section 90 

3.6 Simulation and Modeling in Neurophysiological Research 93 

COMPUTER SYSTEMS LABORATORY 

Summary 99 



Page 

PHYSICAL SCIENCES LABORATORY 

Sunnnary . 105 

Research Project Reports 

5.1 Theor^/ of Bioche:ntcal Separation Technique ■ ■ — 107 

5.2 Theory of the Helix-coil Transformation of Polypeptides 

in Solution ■ -109 

5.3 Molecular Mechanics 111 

5 4 Biophysical Analysis • ■ 113 



s s 



Carbon-13 Nuclear Magnetic Resonance Studies of Peptides and 



13 



Proteins, Including the use of Selective C Enrichment 115 



5.6 Excitation and Transport Properties of Fluids: Laser 
Scattering ■ —117 

5.7 Intermolecular Forces in Biological Structures 119 

5.8 Consulting Ser-'ice? — ■ ■ 122 

5.9 Fundamenfal Studies — 125 

5.10 Nuclear Magnetic Resonance (H4R) Studies of Proteins -127 

5.11 High-Resolution Carbon-13 Fourier Transform NMR Spectroscopy 

at 55 MHz -— ■ ■ 132 

HEURISTICS LABORATORY 

Suiiimary • ■ ■ • 135 

rvesearch Project Reports 

6.1 Automatic Question- Answering and Problem Solving ■ — 137 

6.2 Storage and Retrieval Research ■ 139 

6.3 Automatic Pattern Recognition ■ '141 

6.4 tTLAB 144 

6.5 Information Storage and Retrieval with Index — ■ 146 



ill 



Page 

6.6 Microbiology Data Bank 148 

6.7 Tissue Typing Computer Aids 150 

6.8 Nuclear Magnetic Resonance-Graphics NMR Analysis Program 152 

6.9 Mass Spectrometry Data Interpretation 154 

6.10 Energy State Calculations 157 

6.11 Turnaround Reduction 159 

6.12 Radiation Treatment Planning 161 

6.13 Heuristic Methods for Chemical Structure Searching 163 

6.14 Computer Model of the Human Menstrual Cycle 165 

DATA MANAGEMENT BRANCH 

Summary 167 

Research Project Reports 

Documentation and Systems Support Section 170 

Applied Systems Programming Section 173 

Scientific Application Section I77 

Math /Stat Section ]^80 

Software Support Section ]^84 



Summary of the Associate Director 
July 1, 1972 through June 30, 1973 

The DCRT Scientific and Technical Information Office and the DCRT Library are 
a major NIH resource and a mechanism for communication and education about 
computing. The Division also played a leading role in the NIH Clinical 
Elective for medical students on "Computers in Clinical Medicine." 

THE DCRT LIBRARY is a small collection, maintained primarily for the benefit 
of DCRT personnel and the NIH staff. Some 200 books and 60 reports were 
added during the year, bringing the total of books, reports and bound volumes 
of journals to approximately 2,700. In addition, the Library receives about 
150 periodical subscriptions, with 12 new titles ordered for 1973 and an equal 
number that have been cancelled or ceased publication. Space limitations in 
the Library preclude any attempts to collect everything of interest, but 
definite efforts are made to acquire all material of a high standard of 
excellence and of lasting value in computer science and related fields. 
Members of the DCRT Library Committee assist in the selection. 

During the past year both a new professional librarian and a new non- 
professional library assistant came to the Library. The Library instituted 
a prototype selective dissemination project alerting DCRT staff to articles 
in their areas of interest. The system sends participants the tables of 
contents from five journals of their choice as soon as they arrive. In con- 
junction with the School of Library and Information Services, University of 
Maryland, the Library conducted a user survey questioning DCRT personnel on 
their use of the Library, the kinds of materials they needed, and their 
subject interests. One result of this survey has been a confirmation of 
clinical medicine (as it relates to computers) as an area of interest. 

New Library projects for the coming year include (1) efforts to increase the 
utility of the Library's resources to people outside of DCRT; (2) making 
available bibliographies prepared by the Scientific and Technical Information 
Office; (3) preparing bibliographies for DCRT personnel on selected topics; 
and (4) exploring direct use of MEDLINE as one means of handling requests for 
information about the medical applications of computers. 

THE DCRT SCIENTIFIC AND TECHNICAL INFORMATION OFFICE (STIO) bears overall 
responsibility for communication of scientific and technical information 
within DCRT and between the Division and its environment. The substance of 
this multifaceted role covers three main areas but the expertise has been 
applicable to many other responsibilities of the Division. 

As part of the broad communication of scientific and technical information 
per se during the past year, the office worked on preparation and distri- 
bution of two new DCRT technical reports: Transformation of Observed 
Distributions to Gaussian Form by Eugene Harris and David DeMets, and A 
Structured Assembly Language Source Program Generator by Robert Magnuson. 
The office handles distribution of all previous technical reports, and has 



undertaken development of informative bibliographies on selected computer- 
related topics, deriving from inquiries which it and other parts of the 
Division receive from people looking for an introductory survey on one or 
more topics. 

To facilitate the effective application of computing throughout NIH, the 
office revised and reissued the two basic directories to DCRT resources: 
Services and Facilities of the Division of Computer Research and Technology 
and DCRT Scientific Consultants . The office also began efforts to reach 
more homogeneous groups with information more specifically useful to their 
information processing needs. A fact sheet is under way on different kinds 
of graphic hardware and software at DCRT, as well as an exhibit stand to be 
housed in the corridor between Buildings 12 and 12A. 

The interface between education and communication also became apparent. The 
office helped the Computer Center Branch to reach an audience which had not 
responded to its Interface article describing a seminar on the new Time 
Sharing Option (TSO). The Data Management Branch asked the office to find 
the appropriate government -wide audience for a seminar on its Program 
Generator systems. The CCB seminar had some 90 participants and the DMB 
seminar more than 100. The office continued its role in the development of 
such audiovisual presentations: writing text material in collaboration with 
the DCRT branches and supervising the preparation of visual materials with 
Medical Arts and Photography, DRS. Two successful slide presentations were 
produced for the Data Management Branch. 

The STIO also handles general DCRT public and administrative information 
responsibilities. During the last year visitors and inquiries remained at 
about the same level as the previous year, but some greater interest appeared 
concerning computer-aided diagnosis. The STIO has planned a collaborative 
presentation (with the NIH Clinical Center) to the American Medical Writers 
Association in September 1973. The topic will be "Closing the Gap Between 
Research and Application," the emphasis being on Computers in Biomedicine. 

In support of NIH programs, the STIO prepared four articles for the "Search 
for Health" column, and two radio spots and informational materials for the 
NIH savings bond drive. The NIH Annual Report, Bibliography, Congressional 
Highlights and other such administrative documents continued essentially 
unchanged as background processing for the office during the year, as well 
as the compilation of the DCRT archives, which contain reprints of all papers 
published by the DCRT staff during the year. 

"COMPUTERS IN CLINICAL MEDICINE" was given for the second year as a part of 
the program of NIH Clinical Electives for medical students. This nine-week 
(full time) educational program has emerged as and is now announced as an 
advanced elective for students with prior experience in computing rather than 
an introductory course for students interested in learning something about 
computers. 



Suimaary of the Assistant Director 
July 1, 1972 t hrough June 30, 19 73. 

The Office of the Assistant Director, DCRT, provides three basic capabilities: 
i a) It serves as tne focus for NIti-wide coordination of ADP matters. 

(b) It serves NIH as a central point of contact with the Office of the 
Secretary, other DREW agencies, GSA and 0MB, relative to NIH wide ADP 
policy questions and 

(c) It serves DCRT as a point of reference and coordination to insure 
tnat DCRT's own ADP activities are consistent with NIH, 0/S, GSA and 
0MB policy directions. 

The coordination role is reflected primarily in two activities. All proposals 
for contracts or procurement actions Involving ADP equipment, services, 
programming and design are cleared through this office prior to being 
effected. This provides a continuous opportunity to alert program or 
contract officials to opportunities to reduce costs and/or avoid duplications. 
For example, during the current year, extensive assistance has been provided 
to the Contracting Officer responsible for Frederick Cancer Research Center 
(Fort Detrick) . The contractor is establishing a new ADP capability. This 
office was able to arrange use of inhouse NIH facilities in lieu of more 
expensive separate facilities and assisted in establishing a system of 
project cost accounting for monitoring their $500,000 per year of ADP 
activities . 

A second major undertaking during FY' 73 was providing technical and manage- 
ment leadership for NIH's effort to develop its first comprehensive ADP Plan. 
This plan, now an annual effort, attempts to lay out a two year projection 
for ADP equipment, manpower, contracts and business data systems of both 
the Institutes and NIH's central facility. The greatest value of the 
planning process is that it creates an orderly opportunity for ADP users 
to take stock of their goals and accomplishments. The first annual plan 
co/ered NIH ADP efforts exceeding $23 million per year. 

In serving as a central point cf contact for NIH on ADP related matters with 
DHEW, GSA, OMB, etc., a large number of NIH research and research staff 
members are spared the agony of becoming expert in the many nuances of ADP 
management regulations. Since these regulations are generally written from 
a second generation business data processing point of view, a thorough 
underscanilng of their purpose and operation is necessary in order to 
interpret their application into the NIH research environment. For example, 
in four different cases during the year NIH investigators wanted to expand 
the capability of existing laboratory computer systems. GSA, which had to 
approve the actions, initially took a very narrow "price before performance" 
attitude which was not compatible with the environment in which the systems 
needed to operate. By inviting GSA staff on a tour of our facilities 



Suffice it to say,- the impact of the Pata ^^anagement System stems from the 
ability of a proprammer to create the required computer applications program 
in approximately one-half the time and cost that is involved if traditional 
programming procedures are used. 

The Division's development program is largely centered in the ro^^puter Systems 
Laboratory and focuses primarily on those laboratory and clinical computer 
applications requiring almost instantaneous collection, analysis and -display 
of data as part of carinj^ for a patient or controlling the conduct of an 
experiment. It is the practice of the staff to provide complete systems 
including both equipment and software components for each apolication. This 
practice requires a concentrated, coordinated implementation effort on the 
part of the Laboratory's staff of engineers, mathematicians, computer pro- 
grammers and systems analysts. The staff of this laboratory has been involved 
in the development of more than twenty special purpose, -^'i ni -computer systems 
for NIP scientists during the past few years. The majority of these systems 
have been for laboratory applications. However, the emphasis has shifted to 
clinical applications of computers during the last tv/o years in keeping with 
the DCRT's increasing focus on the Clinical Center. Work on these clinical 
systems is progressing well. 

Particularly noteworthy is the intensive care unit (ICU) for the post-surgical 
monitoring of cardiac patients v/hich has been made availahle to the F'ational 
l!eart and Lung Institute for use and evaluation. If acceptable, this system 
v/ill be expanded to accommodate multiple patients, f^any of the clinical and 
research protocols of the Pulmonary Physiology Laboratory of the I.'HLI have 
been automated v;ith considerable success and additional instruments v/i 1 1 be 
added to the system in this coming 'year. 

A long term commitment of the highest priority has been made this year to the 
support of the clinical laboratories of the Clinical Center. These clinical 
laboratories play a central role in clinical care and clinical research and 
because of their heavy wori<:load require a highly versatile data acquisition 
and reporting system of high reliability. Progress has been made through the 
collaborative efforts of the RCRT and Clinical Center staffs; bov/ever, the 
needs of the clinical laboratories will involve DCRT staff for many months to 
come. 

The research pro-ram of the ^CRl focuses sharply on applied mathematics and 
statistics as utilized in the analysis of biomedical data and is centered in 
the Laboratory of /^oplied Studies, the Laboratory of Physical Science and the 
Heuristics Laboratory. 

Problems of biomedical importance increasingly become more anH more depend- 
ent on the precise and insightful analysis of larpie and variegateH data 
bases. The uni-variate models for clinical and laboratory research are no 
longer applicabl e--the complexity of modern biomedical research requires 
'^ul t i-var iate models involving large volumes of data. As more data are 
gathered, it becomes more necessary to infer and understand patterns 
in the data, to categorize the data into classes and to model the under- 
lying biological and physical systems v;hich are generating the data. The 
niethodologi es for achieving these ends are based squarely on the intellec- 
tual procedurrs of deterministic, stochastic and pattern recopnition or 



heuristic mathematics. Exploitation of these mathematical models and pro- 
cedures is made practical by the existence of the computer and the computer 
software which al1ov;s rapid and accurate logical evaluation of data. The 
staffs of the three laboratories mentioned above have provided an extensive 
array of computer-based software systems and mathematical techniques to the 
MIH scientists for data analysis. 

A representative but hardly exhaustive list of these research activities serves 
to illustrate the point. Scientists in the Laboratory of Applied Studies have 
formed a very firm partnership with some of the MIH clinical staff. The impor- 
tant yield of this collaboration has been the critical evaluation of computer 
programs v;hich analyze the electrocardiogram, a auantitative basis for the 
analysis and interpretations of "normal variation" in the blood chemistries 
and an automated and a auantitative basis for the analysis of nuclear isotope 
data used in diagnosis of disease and the description of the dynamic function 
of oreans. It is fair to say that the clinical research conducted by f'JH 
scientists will benefit considerably from these newly available intellectual 
tools for generating and analyzing data. 

The staff of the Heuristics Laboratory has undertai<en very difficult bio- 
medical problems with considerable success in some instances. It is now clear 
that mass spectrometr ic data are uniquely productive in both the laboratory 
and the clinic for identifying chemical entities v/hich are important biological 
moieties or are toxic agents to the human. A computer-based library of mass 
spectrometr ic data for a large number of chemicals plus the programs for 
interrogating this data base have been made available by the staff. The 
system has been used for diagnostic, toxicological and educational purposes on 
an international as v/ell as national basis. An adjunct to this identification 
system are the new mathematical procedures which serve to correlate the mass 
spectrometr ic data with chemical structure and in turn, with biological func- 
tion. 

A very challenging study undertal<en by the staff is concerned with the pro- 
cedure of tissue typing, the results of which presumably predict the fate of 
a tissue or organ transplant. Little progress has been made by either manual 
or mathematical procedures, largely because the antigen typing materials are not 
specific for single antigens. The point here is that the manual interpretation 
of the typing data is grossly inadequate and the hope is that a formal mathe- 
matical procedure can deal with the complex data in terms of distinguishing 
insightful patterns. The existence of the DCRT staff makes such a study possi- 
ble; it is hoped that the mathematical approach will yield greater insight to 
this important and difficult problem. 

The Physical Sciences Laboratory has continued to emphasize the development of 
the auantitative and theoretical basis of procedures such as ul tracentr i fuga- 
tion, chromatography, electrophoresis and nuclear magnetic resonance spectros- 
copy, that are used for the physical chemical characterization of macromol ecul es. 
The goal is the development of better numerical analytic methods for the more 
accurate processing of the physical chemical data that arise from these methods. 
Also, considerable progress was made in developing quantitative methods for 
study of conformational structure of biological polymers. Particularly note- 
worthy was the continued progress made in the quantitative elucidation of the 



electrostatic (Coulombic) and el ectrodynanic (van Her Waals) forces that play 
such an important role in determining chemical and ^>ioIogical structure. Tom- 
pleaienting these quantitative studies of rrolecular structure has been the 
development of new PDP-10 computer-based hardware and software v/hich greatly 
facilitates the graphic presentation of data whereby the atomic structure and 
electron density data may be matched to yield precise information on the 
nolecular and conformational structure of biologically imoortant macromol ecul es . 

An obvious conclusion must be drawn fro.T this report which is contrary to the 
existing notion that satisfaction is derived solely with the identification of 
a human with an important end-result. On the contrary, an enormous sense of 
pride and satisfaction accrues to the engineer, mathematician and computer 
scientist who provides the capability for ouality performance of the scientists 
in the "^'IH setting. 

It is clear that a medical information processing system has been developed 
in support of the conduct and management of the diverse research programs of 
the fiational Institutes of Health, ••'any individual clinical and laboratory 
applications have come under development during the five years of the Division's 
formal existence. In general the emphasis has been focused on the data acqui- 
sition and data analysis needs which characterize the research envi ronm.ent . 
Central to these requirements has been the development of a mul t imachi ne computer 
network organized around a large central facility which is itself a network of 
machines and augmented by linked peripherally located computers, each dedicated 
to a special task or set of tasks. 

For file processing, a general purpose data management system has been created 
based on the concept of a pre-processor . The macro-processor is a recursive 
string manipulator which accepts macro statements as input and creates com- 
piler source statements as output. These source statements are compiled and 
executed as a program. The advantages of the macro-processor arp a greatly 
increased degree of machine independence and a considerable reduction in the 
time required for program developmient. The macro statements are considerably 
closer to user-oriented discourse than are standard programming language 
source statements. The macro statement does generate a comiplete set of source 
statements for file creation, maintenance, and retrieval. In addition, the 
full power of the compiler is made available for f i 1 e orocessi n? tasks. 

For the data analysis requirements, a large number of numerical analysis 
library routines have been prepared and made available on a program-call basis 
under a computer-based system that provides interactive terminal support to 
the user. Considerable attention has been ^ocused on the computer-leased 
processing of natural language data since the majority of medical data are 
foripulated in natural or ordinary language. Currently pathology diagnoses 
in natural language form are automatically indexed by computer programs v;ithin 
the machine and organized into formatted files suitable for precise retrieval of 
both medical documents as well as explicit items of medical information. 

These computing tools, namely the computer network, the data management system, 
the computer-hased data analysis capability, are toeing t^lended together into a 
medical information system for the research environment. * 



Summary of the Associate Director 
July 1, 1972 through June 30, 1973 

The DCPT Scientific and Technical Information Office and the DCRT Library are 
a major NTH resource and a mechanism for communication and education about 
computing. The Division also played a leading role in the NTH Clinical 
Elective for medical students on "Computers in Clinical Medicine." 

THE DCRT LIBRARY is a small collection, maintained primarily for the benefit 
of DCRT personnel and the NIH staff. Some 200 books and 60 reports were 
added during the year, bringing the total of books, reports and bound volumes 
of journals to approximately 2,700. In addition, the Library receives about 
150 periodical subscriptions, with 12 new titles ordered for 1973 and an equal 
number that have been cancelled or ceased publication. Space limitations in 
the Library preclude any attempts to collect everything of interest, but 
definite efforts are made to acquire all material of a high standard of 
excellence and of lasting value in computer science and related fields. 
Members of the DCRT Library Committee assist in the selection. 

During the past year both a new professional librarian and a new non- 
professional library assistant came to the Library. The Library instituted 
a prototype selective dissemination project alerting DCRT staff to articles 
in their areas of interest. The system sends participants the tables of 
contents from five journals of their choice as soon as they arrive. In con- 
junction with the School of Library and Information Services, University of 
Maryland, the Library conducted a user survey questioning DCRT personnel on 
their use of the Library, the kinds of materials they needed, and their 
subject interests. One result of this survey has been a confirmation of 
clinical medicine (as it relates to computers) as an area of interest. 

New Library projects for the coming year include (1) efforts to increase the 
utility of the Library's resources to people outside of DCRT; (2) making 
available bibliographies prepared by the Scientific and Technical Information 
Office; (3) preparing bibliographies for DCRT personnel on selected topics; 
and (4) exploring direct use of MEDLINE as one means of handling requests for 
information about the medical applications of computers. 

THE DCRT SCIENTIFIC AI^ID TECHNICAL INFORMATION OFFICE (STIO) bears overall 
responsibility for communication of scientific and technical information 
within DCRT and between the Division and its environment. The substance of 
this multifaceted role covers three main areas but the expertise has been 
applicable to many other responsibilities of the Division. 

As part of the broad communication of scientific and technical information 
per se during the past year, the office worked on preparation and distri- 
bution of two new DCRT technical reports: Transformation of Observed 
Distributions to Gaussian Form by Eugene Harris and David DeMsts, and A 
Structured Assembly Language Source Program Generator by Robert Magnuson. 
The office handles distribution of all previous technical reports, and has 



undertaken development of informative bibliographies on selected computer- 
related topics, deriving from inquiries which it and other parts of the 
Division receive from people looking for an introductory survey on one or 
more topics. 

To facilitate the effective application of computing throughout NIH, the 
office revised and reissued the two basic directories to DCRT resources: 
Services and Facilities of the Division of Computer Research and Technology 
and DCRT Scientific Consultants . The office also began efforts to reach 
more homogeneous groups with information more specifically useful to their 
information processing needs. A fact sheet is under way on different kinds 
of graphic hardware and software at DCRT, as well as an exhibit stand to be 
housed in the corridor between Buildings 12 and 12A. 

The interface between education and comirrunication also became apparent. The 
office helped the Computer Center Branch to reach an audience which had not 
responded to its Interface article describing a seminar on the new Time 
Sharing Option (TSO) . The Data Management Branch asked the office to find 
the appropriate government-wide audience for a seminar on its Program 
Generator systems. The CCB seminar had some 90 participants and the DMB 
seminar more than 100. The office continued its role in the development of 
such audiovisual presentations: writing text material in collaboration with 
the DCRT branches and supervising the preparation of visual materials with 
Medical Arts and Photography, DR3. Two successful slide presentations were 
produced for the Data Management Branch. 

The STIO also handles general DCRT public and administrative information 
responsibilities. During the last year visitors and inquiries remained at 
about the same level as the previous year, but some greater interest appeared 
concerning computer-aided diagnosis. The STIO has planned a collaborative 
presentation (with the NIH Clinical Center) to the American Medical Writers 
Association in September 1973. The topic will be "Closing the Gap Between 
Research and Application," the emphasis being on Computers in Biomedicine. 

In support of NIH programs, the STIO prepared four articles for the "Search 
for Health" column, and two radio spots and informational materials for the 
NIH savings bond drive. The NIH Annual Report, Bibliography, Congressional 
Highlights and other such administrative documents continued essentially 
unchanged as background processing for the office during the year, as well 
as the compilation of the DCRT archives, which contain reprints of all papers 
published by the DCRT staff during the year. 

"COMPUTERS IN CLINICAL MEDICINE" was given for the second year as a part of 
the program of NIH Clinical Electives for medical students. This nine-week 
(full time) educational program has emerged as and is now announced as an 
advanced elective for students with prior experience in computing rather than 
an introductory course for students interested in learning something about 
computers. 



Summ ary of t he Assistant Director 
July 1, 19 72 through J u ne 30, 19 73. 

i'iic Office ot the Assistant Director, DCRT, provides three basic capabilities: 

t'a) It serves as the tccus tor NIri-wide coordination of ADP matters. 

(,d) It serves NIH as a central point of contact with the Office of the 
Secretary, other DHEW agencies, GSA and 0MB, relative to NIH wide ADP 
policy questions and 

(c) It serves DCRT as a point of reference and coordination to insure 
tnat DCRT's own ADF activities are consistent with NIH, 0/S, GSA and 
OMd policy directions. 

The coordination role is reflected primarily in two activities. All proposals 
for contracts or procurement actions involving ADP equipment, services, 
programming and design are cleared through this office prior to being 
effected. This provides a continuous opportunity to alert program or 
contract officials to opportunities to reduce costs and/or avoid duplications. 
For example, during the current year, extensive assistance has been provided 
to the Contracting Officer responsible for Frederick Cancer Research Center 
(Fort Detrick) . The contractor is establishing a new ADP capability. This 
office was able to arrange use of inhouse NIH facilities in lieu of more 
expensive separate facilities an'! assisted in establishing a system of 
project cost accounting for mcnltoring their $500,000 per year of ADP 
activities . 

A second major undertaking during FY' 73 was providing technical and manage- 
ment leadership for NIH's effort to develop its first comprehensive ADP Plan. 
This plan, nov; an annual effort, attempts to lay out a two year projection 
for ADP equipment, manpower, contracts and business data systems of both 
the Institutes and NIH's central facility. The greatest value of the 
planning process is that it creates an orderly opportunity for ADP users 
to take stock of their goals and accomplishments. The first annual plan 
c'--.'ered Nlri ADP efforts exceeding $23 million per year. 

In serving as a central point of contact for NIH on ADP related matters with 
DHEW, GSA, 0MB, etc., a large nuro.ber of NIH research and research staff 
members are spared the agon3/ of becoming expert in the many nuances of ADP 
management regulations. Since these regulations are generally written from 
a second generation business data processing point of view, a thorough 
understanding of their purpose and operation is necessary in order to 
interpret their application into the NIH research environment. For example, 
:n four different cases during the year NIH investigators wanted to expand 
the capability of existing laboratory computer systems. GSA, which had to 
approve the actions, initially took a very narrow "price before-performance" 
attitude which was not compatible with the environment in which the systems 
needed to operate. By inviting GSA staff on a tour of our facilities 



we were able to convince them to allow us to proceed. In other cases an 

inordinate amount of time is expended with less success in attempting to 

reconcile research needs with the kinds of business rules used by GSA and 
0MB. 

This office's coordination role with DHEW also provides an opportunity 
for NIH to influence DHEW policy. The present ADP planning system was 
conceived by a task group headed by the NIH member. We were thus able to 
keep the Office of the Secretary's attention prim.arily focussed on 
management uses of ADP. This left NIH research programs largely free from 
intervention by the Office of the Secretary. Also NIH is the first agency 
in DHEW to automate its ADP equipment inventory reporting system. This 
is now being- eopied by several other agencies and attests to NIH's leadership 
in using ADP for tasks associated with ADP management. 

As regards DCRT's own operation, the Office of the Assistant Director 
coordinates all ADP procurements, contracts and work orders which have a 
technical or engineering impact on our computer operations. By bringing an 
understanding of technical considerations into these matters, better and more 
responsive services are obtained from the central service groups. Further, 
much DCRT staff time is saved and can be applied to technical problems 
rather than to following up on administrative details. 



Serial ilo , 

1. Office of the Dii-r-ctor 

2. 

3. r.et^es 'a 

PnS - '!!■' 

In 1 i V i '.'ual Project Report 

.July 1, 1972 throu?;h Juno 30, 1073 

Project Title: Visual and oiolc-ical Shane 

Principal Investigator: f'arry Blun 

Previous Serial ilunber: DCRT 1.1 

Other Invest ii^ators : Virr^il Carlson (f'l.P) 

Franc ine Froie ('l!"i^' -^ U. o^ f'H.) 
Judith Prev/itt C^'^PT) 
nichard L. Mehher Cun^) 

■^ooperat in^; Unit: 'llf'H, Lab. of Psycholo-iy, 

Section on Percnotion. 
DCRT, Data i iana^e'ient Branch, 
;]ICn, Oral fledicine an'^ Suri^ery C ranch 

ilan Years 



Total : 


0. 


, 9 


Profess ional : 
Other: 


0, 

0. 


, J 
,0 



Project cle scr i pt ion : 

Obj ec t i ves : 

The overall objective of this project is to ieveloo a fornal 
Jescription for bioloc^ical shapes and their 'evelopient, and 
apply these to (1) the description of cells, or-;ans and 
ori^anlsnis, and (2) the understand I n~ of the visual process, 
both psychologically and neurophys I ol o;; i ca 1 1 y . 

f'.ethods Znployed: 

Continue to fornal Ize and exten' the s^eonetry of biolo<Tical 
shape based on t\\e i;ro'./th prinitive. Txt^n' an-' use both 
iiand and connuter -lethods for extracting these descriptions. 
Apply then to (1) cells and tissues talcen fron blad-'er 
sections for cancer patholor;y analysis and (2) skeletal 
description and ';ro\/th usini the cranial cavity ta';en fron 
dental skull radi o;;raphs . Continue exploration of visual 
shape process by revlev/in^ relevant experinental areas in 
i)sychol o<:y and neurophysiology. Desl-^.n and carry out 
experiments for refutin^^ or valldatins; the visual theory. 

9 



"■ajcr F in--' in 15: 

Tecimiqjes for the aato^iatic analysis by co muter of new 
shape oroperties have been dovelooei ani are cont i tu i n-^ . \ 
psychophysical shape experi lent out into serjojs -inubt the 
nost accente'^ theory for coiin-^ of curvp ' for^is ^n' ./as lore 
co'ipatible with the iiereiri rirooosei t^'eory. \ n^..' set of 
exoeri lents are bein^ un-lertal'-pn to coipare it v/ith f^o-^ 
other DOSS i hie theories. 



S i -^n i f i cance to Fi io leii ca 1 Research 



an 



the '^rn;ra~i of nrrj 



At the 


lost direct 


teciin ici 


ues necessar 


Id ter ia 


1 (s 1 i :lus, r 


theory 


cot! t r i bu tos 


J lol 0^ i 


cal areas. 


for ' !e s 


c r I p t i n v: 


■jo n-^e ' 


(?■'- I'i^r taxon 


a n ^! a J t 


la t ic . CpC' 


le rcept 


ion ■, ' 1 1 i c ^1 is 


process 


and lia^.nos 


i nfor"ia t ion . Th i r ' 


rie rvo>.i-3 


sysLen or -a 


ne jroo'T 


y s i 1 •^ y '') f 


ti'.c C'io 


Th.is exte 


:jtnsory 


or,-;an iza t i u 


e uryon 


ic an.! later 


Liiooret 


i ca 1 pr i nc i p 


biolor^ical sha.^e. 


shown t 


be i nvc rs i 


wor!;, i 


f correct, Ir 


i"! porta 


nt areas of 


i nves t i 


":a t i on be for 



level, tlie ./o 
y for the aut 
a ! i o^rn phs, e 
a f jn -!a len ta 1 
First, it pro 
ce 1 Is, or '^ans 

1 i c an i -' i a ^ 
on ', it onens 

nee 'e- f^^r 11 
tic cont r i ivj t 
, it proposes 
n i z a t i o n . ; h i c 
V i s i o r; a n ■ ■ to 
PS i on en CO 1 oa 
n. Fojrtii, i 
, ^ro-..'th, an! 

1 e s apply to 
The visual an 
ons eq ja 1 1 y .; 
as potent i a 1 
biology '.;h i cti 



rb -ia';es available ne\/ 
oiatic analysis of nictorial 
tc.). Far beyrm-' tliat, the 

in s i "^lit i a to seve ra 1 
poses a ■ lo rnhol OT i -q 1 /;.ase 

an! or '^an is -IS. This './ill 
nostic o.iroMses, br^th huian 

ur-) the area n^ sbaoe 
n'erstan^'I'-'': tlie "ea'^in^ 
ions f r o n v i s ' : r 1 

ne./ orinciolcs oF central 



h -lay be i loortant to tl 
"e^eral 'in '^r :", top 'i n-^ - 
sses ■lot'^r, as '..'r'Tl as 
t has inolicati'^ns in 

ievo 1 opoen t since the 
the efficient co'lin-^ of 

' the -;ro\ .'th coii-v!; are 
ell t re ate-!. Thus, the 
for o;ienio" ji) neu' an"! 
Iiave been resistant to 



Proposed '"our se: 

Theoretical v;or'-'. will continue. '\x;)p r i letna 1 :irir'.: ^n sh.ane 
psychoohys i cs ./ill continue. It is hone ^ that experin^nts 
in visual physiolor^.y v/i 1 1 be started. Pol 1 aborat i ve './orl; 
for the application of this descriotion to cell nuclei! and 
skull ,;rowth is !'e!n; exploro-i. 



loners an 



\iBr . 



lone 



PUbl i ca t i ons : 

riun, f. : riolor;ical shape an-* visual science (^art I) 
.!. Theor . Tiol . ZZ: 2')5-2?7 , 1373 

Author: Froie, F. S.: ^ "'sychonhys i ca 1 .'Jtudy of Ghape 
M i r^.nnent . Coll ere Parle, 'id. i^ooouter Science '"poter, 
Lnlversity of '.'A., 1372, U2po. 



10 



Serial No. 1.2 

1. Office of the Director 

2. Medical Information Science 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1972 through June 30, 1973 



Project Title: Clinical Center Admissions 

PrevioL's Serial Nitraber: None 

Princip?.! Investigator: William C. White 

Other IriT'astigators: Marvin Katz 



Cooperatiag Units; 



Office of Clinical and Management Systems, Clinical 
Center 



Man Year?; 




Total : 


0.5 


Frcf esslonal ; 


0.5 


Other: 


0.0 



Project Description: 

Qbjr-ctives: 

To iievelop and maintain an on-line system for collecting and editing Clini- 
cal Center admissions data and establishing a file of biographic data on 

Clinical Center patients. 

Background : 

Computer Admission Systems have been designed for the Clinical Center in 
£h« past, however none of these systems was on-line and interactive to 
permit coi'ip?Jt£r editing and validation of the data as entered. The avail- 
ability of a commercial program for interactive terminal transactions on 
the central computer facility prompted this latest admissions system design. 

Methods Employed: 

Use of the "Customer Information Control System" (CICS) allows the use of 
the PL/i prograjuming language as well as Assembly language programs. The 
progrsjES were tested and modified to fit the requirements and convenience 
of the potential users in the Clinical Center. A registry file was estab- 
lished including all machine readable information on all past Clinical 



11 



Center admissions. This registry file with its Soundex equivalent pro- 
vides the key to the retrieval of Biographic and Admission data and 
future patient medical record files. 

The Admissions System has been in use for more than a year with three 
terminals at the Admissions Office in the Clinical Center. Admissions 
personnel have been instructed and have been using the terminal for 
collecting and printing admissions data on a 16 hours a day, 6 days a 
week, schedule since October. One terminal has been installed in the 
Medical Records Department for collecting discharge data. 

The Admissions System has been expanded to include information about 
patient transfers and discharges. Several useful computer output lists 
are available, such as the patient census and discharge summaries. 

Significance: ^ . . 

The Admissions System provides the base for the development of a computer 
stored Patient Record for use in patient care and clinical research. 

i 
Proposed Course: 

The start of an on-line patient record system will be initiated using the 
doctor's orders and drug administration data for a test system. This 
system will be based on a single nursing unit as a program module to allow 
easy adaptation to the requirements of any particular area. 



12 



Serial No. 1.3 

1. Office of the Director 

2. 

3 . Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1972 through June 30, 1973 



Project Title; Computer Acquisition of Physician Orders 

Previous Serial Number: Same 

Principal Investigator: Robert D. Gilbert, M.D. 

Other Investigators: William C. White, Marvin Katz, Linda McNeely 

Cooperating Units: National Institute of Arthritis and Metabolic 

Disease, Arthritis and Rheumatism Branch, 
Department of Nursing, Clinical Center 

Man Years: 

Total: . 2.5 

Professional: 2.0 
Other: 0.5 

Project Description: 

Objectives: 

The ultimate objective of this project is to acquire in machine processable 
form those physician orders which determine and initiate essentially all of 
the diagnostic and therapeutic activities of the hospitalized patient. In 
particular, those orders pertaining to drug therapy are of interest. Such 
information provides the basis for ultimate analysis of drug efficacy, 
toxicity, and interaction, as well as the eventual reduction of drug 
administration errors. 

Methods Employed: 

A DCRT supported public text editing system called WYLBUR has been the 
means used to acquire the basic data. Data is entered via computer type- 
writer terminal at the nursing unit, and is transmitted by telephone to 
the DCRT central computing facility for batch processing. 

For on-line entry of data, programs are being developed to utilize the 
"CICS" computer system and access the Computer Admission Systems (see 
Serial 1.2). 



13 



Major Findings: 

The batch processing system has been in full operation on a single nursing 
unit for over two years. Nursing response to the system has been unani- 
mously favorable. At least seven documents of clinical use to nurses are 
being generated daily from input data. Achieving compatibility with 
nursing has been the initial objective of the system, and this has been 
accomplished. Costs have been reduced by consolidating programs and 
executing the compiled and stored run modules. 

Proposed Course: 

Additional programming is underway to allow on-line, interactive entry and 
editing of the doctor's orders and other pertinent patient data for use on 
the nursing unit. These programs will store the information for ready 
display of the current status of a patient, his active orders and the 
medications administered or due to be given at a specified time. The 
programs will run under the "Customer Information Control System" (CICS) 
on the IBM 360/370 computers at the central computer facility and will 
link to and become a part of the Clinical Center Admissions System. 

Honors and Awards: None 

Publications : None 



14 



Serial No. 1.4 

1. Office of the Director 

2. Medical Information Science 

3. Bethesda 



PHS - NIK 
Indiviciual Project Report 
July 1, 19 71 through June 30, 19 72 



Froject Title: 
Previous Serial Nvmber; 
Principal InvesLigator: 
Other Investigators: 
Man Years: 



AuciTOiated Proceiising of Medical Language 

4.9 

Mil OS Pacsk 

George Dunham; William C, White 



Total! 2,5 

Professional: 2,5 
Other: 0.0 

Project Description: 

Objectives ; 

Tne najor objective at the project was design of a new and more sophisti- 
cated system for automated prucessing of medical language based on the 
experience with the praaant SNOP encoder, and medical data base. 

Sackgiound: 

Fcr the past several years, an effort has been underway at the DCRT to 
develop an automated information retrieval system to handle pathology data 
which were manually encoded and made available to the medical scientist 
for research purposes „ We have developed a system for automated encoding 
(indexing) which became operational several years ago. The system includes 
the acquisition of textual information, the interrogation of a dictionary 
\3H0P) , a set of raorphosyntactic rules and a set of syntactic-semantic 
rules which are necessary for the identification of the information con- 
tent of the input messages and their subsequent encoding. 

Progress during FY 73: 

As the encoder became operational, refinements to the program and the dic- 
tionary were continued. Encoding of the Clinical Center surgical patho- 
logy diagnoses for 1970 is complete and for 1971 almost complete. 



15 



We have completed a generalized computer-oriented system for morphosyn- 
tactic analysis of medical English, the components of which ar-^i as fol- 
lows: 

1. The assignment of syntactic markers to word forms on the basis of 
their suffixes. 

2. Transformation rules which include adjective-to-noun, noun- to- 
adjective, noun plural- to- noun singular and -ED and -ING forms to 
nominal form transformations. 

3. Assignment of semantic markers to a selected list of suffixes. 

We have also continued our work on the refinement of morphosemantic anal- 
ysis of compound word forms. This work is closely connected with the 
development of procedures for the construction of a root medical lexicon 
to be used in computer processing of medical language. A man-m.achine 
interactive lexical refinement algorithm was developed in this direction. 

Proposed Course: 

1. Encoding of pathology diagnoses for the years prior to 1970 in coop- 
eration with the Laboratory of Pathology of Clinical Center, 

2. Construction of medical stem dictionaries in English, German and 
French, 

3. Formalization of procedures for the semantic interpretation of medical 
records. 

Publications: 

1. Pacak, M. , Cousineau, L. , White, W.: The Segmentation Approach to 
Dictionary Construction, Journal of Clinical Computing , Vol. 2, No, 3, 
November 1972. 



16 



Serial No. 1. 5 

1. Office of the Director 

2. Medical Information Systems 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1972 through June 30, 1973 



An Information Processing System for Pathology Data 

1.6 

Martin Epstein 

Samuel Harper 
Paul Kalkowski 

NCI, Laboratory of Pathology 



Project Title 
Previous Serial Number 
Principal Investigator 
Other Investigators 

Cooperating Unit 

Man Years: 

Total : 2.3 
Professional: 2,4 
Other : .4 

Project Description: 



Objectives: 

The objective of this project is to develop and maintain an information 
processing system for pathology data. This project includes the creation 
and updating of a data base for Autopsy, Surgical Pathology, and Cytopath- 
ology data, and retrieval and reporting from this data base through queries 
submitted by medical investigators. The Systematized Nomenclature of 
Pathology (SNOP) is maintained on disk for use in automated encoding and 
in the development of suitable encoding dictionaries for computer process- 
ing. Techniques are being developed to classify automatically medical 
diagnoses selected on retrieval from data base. 

Progress: 

During the past year work continued on the expansion of the surgical 
pathology data base to include reports for the years 1968 through 1973. 
The reports were keyboarded at a terminal using the WYLBUR text editing 
system. The surgical pathology reports for 1970 and 1971 were automati- 
cally encoded, a data file created and retrievals performed against the 
file. The input to the current data base is post-edited by a staff 
pathologist who suggests modifications to the SNOP dictionary to improve 
the encoding accuracy of subsequent encoder runs. 



17 



■■■„,jCT=minii«aa 



Several collaborative studies with U.S. and foreign investigators have 
been initiated to determine the adequacy and relevance of the information 
handling and linguistic procedures that have been developed to process 
surgical pathology diagnostic statements and related data. The availability 
of data from other institutions has served to verify the utility of the 
research pursued to date in processing surgical pathology data. 

Proposed Course: 

Extension and modification of the currently operational information 
processing system for pathology data. Work is continuing on: (1) the 
design and development of encoding dictionaries to organize and categorize 
medical information for storage in a data base and subsequent use in 
question-answering situations, (2) the definition and description of 
efficient data and storage structures needed to represent medical data, and 
(3) the general problem of language understanding in a machine and the 
high level languages needed to achieve this understanding. 

An effort is underway to design and implement a data acquisition and 
management system to process all the reports generated in the Department 
of Pathology. Such a data management system will provide for rapid and 
efficient processing of medical reports within the department and provide 
verified input to the computer systems under development. 



18 



Juxy 1, 1972 thru June 30, 1973 

PUBLIC HEALTH SERv^XCE-NATIONAL INSTITUTES OF HEALTH 
DIVISION OF CO'«IPUTER RESEARCH AND TECHNOLOGY 

1. DCRT--2 

Serial Number 
'.OWFUTER CENTER BRANCH 3« J. D. Naughton 



Branch Chxef 



I . SUI^iMARY 



The Computer Center Branch designs, implements and operates a 
large geaeral-purpose computer utility to n&et laost effectively 
the dynamic and diverse requirements of both N.I.H. research 
investigstcrs and rrianagers in the support of modern medicine. 
This charge includes the original development of new system 
facilities to meet the unique requirements of the NIH mission in 
order to bring the full power of the computer to bear on problems 
at every level of research xn many diverse locations. The core 
of this computer facility is a network of computers and remotely 
located terminals, which, by means of modern communications 
techniques, extends the pov;er of the computer directly into 
laboratories and offices throughout N,I.H. This provides 
imiriediate access to the computer at the source of the problem 
thus eliminating the delay in the research program and making 
more efficent use of critical manpower than more traditional 
methods. An inherent responsibility of the Computer Center is 
the continued research ir i-.o r jw methods of extending the network 
evQTi further into the research environment while continually 
adapting to the constant impact of new knowledge and program 
direction. 

A fvill spectrum' of computational services are provided to all 
Institutes and Divisions of the NIH on a fee~for-service (cost 
recovery) basis. These facilities include conversational pro" 
graEttP.ing. graphics, microfilm output, text editing, remote job 
:'nt~Yy time sharing and batch processing. Large system^s as well 
- iTiini-computers and terminals are lashed together providing 
-istributsd capacity" available at many levels. Research into 
ths computer and infonriation sciences coupled with close coop- 
aration between the N.I.H. medical investigators and the computer 
scientist have incorporated computers directly into the research 
environment where they can perform most effectively in attacking 
the complex problems of modern medicine. 

The medical research programs of N.I.H. require the most powerful 
and flexible of computer services and tools available today. The 
computer network provided must have a distributive power that is 
easily accessible on demand to scientists in the laboratory 
itself. The goal is to mold^ polish and, in general, enhance the 
computer into a complete tool for medical research and its 



19 



administrative support. New areas of computer applications will 
be sought out continuously, in conjunction with a comprehensive 
program of educating new and old computer users in how to use 
computers most effectively for their mission. 

1973 Activities 

The activities of the Computer Center are directed at providing 
the best computing environment and the widest range of computing 
tools, that will satisfy all the facets of the N.I.H., 
scientists, researchers as well as their administrative support. 
The success of these activities and the Computer Center is best 
depicted by the cold statistics that show the continually 
increasing demand for all the center's services by the N.I.H. 
community. The overall workload measured in number of jobs 
processed increased during 1973 by more than 25%, from 105,000 
jobs to over 140,000 jobs per month. 

The full effectiveness of the Computer Center telecommunications 
network was recognized fully during 1973 when over 90% of all 
jobs submitted for processing arrived by telephone line. The 
number of interactive typewriter terminals located in labo- 
ratories and offices increased by over 200 for a total in 
excess of 600. Coupled with this was an increase in the number 
of daily terminal sessions from 1500 a day to over 190 a day. 
The increasing effectiveness of remote batch terminals is evident 
as the number more than doubled from 27 to 50. Over 35,000 jobs 
a month now flow through these terminals. This ability to extend 
the full power of the computer directly to the source of the 
problem has made it possible to apply modern computational 
methods to many areas where it was not feasible in the past. 

Despite this tremendous increase in the amount of work submitted 
for processing, the users of the Computer Center continue to 
receive exceptionally rapid turnaround time for all computational 
services. To maintain the responsiveness of the system the 
computer configuration supporting this workload was continually 
changing to provide more compute power or new computer tools or 
techniques. During 1973 the 360/50 was replaced by a 370/145 and 
a 360/65 was replaced by a 370/165 which increased the compu- 
tational capability by a factor of 4 while reducing operating 
costs slightly. This move not only increased power but enables 
the Computer Center to test the latest in computer techniques 
that will eventually be passed on to the users in the form of new 
computational facilities. 

The increasing volume of work processed had a direct cost benefit 
in that it allowed the Computer Center to offer a 20% reduction 
in computer rates for the overnight work class. In addition, 
improvements on the internal efficiency of the system made by the 
Computer Center allowed other operational changes benefitting 
users. Among these was an increase of 50% in the available core 



20 



for Class B, C and G jobs and an increase of 1/3 for Class H. 
The available CPU time for all classes of work was likewise 
increased significantly, 33% for Class A and 20% for Class B. 

Many new computational services were offered to the N.I.H. 
community this year. Most significant was the announcement of 
the timesharing option (TSO) of the operating system. This 
system offers conversational debugging and more importantly the 
capability for the user to construct his own interactive system 
in any of four standard computer languages. The TSO, CPS, 
WYLBUR, CICS combination now give the users a complete selection 
of interactive systems to match the requirements of any job. The 
number of simultaneous CPS users increased to 38 while WYLBUR's 
simultaneous users increased to 184 and continues to grow. The 
attraction of these terminal systems should be enhanced further 
by the planned announcement of fast CRT support for all terminal 
systems. Microfilm was offered as a new output medium for 
computer users. The microfilm approach provides a vehicle for 
not only cheaper but faster and easier handling of voluminous 
reports. The use of microfiche with titling and automatic 
indexing enables users to set up retrieval systems for various 
data bases. These microfiche retrieval systems enable the user 
to have immediate access to large data bases without the cost of 
storing them on-line. 

Among the new tools offered users was the announcement that the 
PDP-10 computer's timesharing system would be on a fee-for- 
service cost recovery basis. This system offers many new 
facilities in support of laboratory research. One of the most 
outstanding being a wide array of on-line graphics hardware 
capability, coupled with supporting languages such as the 
3-dimension graphics language. 

The macro molecular display and manipulation program was 
completely revised, utilizing the existing hardware better to 
allow the handling of larger molecules. The plotting 
capabilities of the system were vastly improved as was the 
generation of molecular conformations for energy calculations. 
The system will be further enhanced by the addition of a surface 
drawing display system. This macro display program was a vital 
part in the development of computer representation and 
manipulation of macro molecular electron density surfaces. 
Starting with an electron density map crystallographers can 
construct and refine the atomic coordinate model completely 
within the computer. Several macro molecules are in the process 
of structure determination presently. These techniques will 
reduce significantly the professional manpower needed for 
molecular structure project. 

The art of Remote Computer Graphics was advanced with the 
implementation of the Omnigraph system. This allowed scientists 
to develop interactive graphics systems in a choice of languages 



21 



IHHGBi^iiBilBiBHil 



and to display their results in a variety of CRT display 
terminals. The Cambridge (England) Crystal File containing all 
the bibliographic and structure information of all crystals 
catalogued since 1960 was made available to the N.I.H. research 
community. This file is an invaluable reference tool for all 
crystallographers at N.I.H. 

Close contact between the staff and users of the Computer Center 
continues. Although the size of the user community has grown 
quite large and the widespread use of telecommunications prevents 
personal contact with each individual, the user community remains 
current on recent developments in computing at N.I.H. thru 
INTERFACE and the Computer Center Users Guide. The User's Guide 
was kept current with timely updates and INTERFACE continues to 
broadcast the latest news in computing to and beyond the entire 
computing community. INTERFACE is almost as widely read by those 
outside the N.I.H. user community as by those inside. This 
attests to the quality of the publication. The data-line portion 
of INTERFACE serves an important function, as many scientists, 
and administrators have made invaluable suggestions thru this 
mechanism. Thru data-line the user can express his opinion or 
offer his suggestion on how the Center can best meet his needs. 
The suggestions offered and opinions spoken are given much 
consideration and help the Center to mold their future course of 
action. 

The accomplishments of the Computer Center are only as good as 
the use that the N.I.H. scientist makes of the system. The user 
makes the best use of the computer utility when he understands 
what he is using and how to use it most effectively for his own 
particular needs. The Computer Center continued to improve its 
training program by offering over 80 courses to over 2,000 
scientists,, administrators and support personnel during 1973. 

1974 Plans 

The Computer Center looks forward to the opportunity to provide 
new computational capabilities in support of the N.I.H, mission 
next year and the years beyond. Plans have been made both for 
improving the effectiveness of existing computational services 
and for adding new functions. In addition to increasing CPU 
capacity, new virtual system software and hardware is planned to 
give users more interaction and direct control of their projects. 
A new SUPERWYLBUR providing totally new functions not available 
in the present system will be installed early in 1974. The 
present PDP-10 system will be replaced by a larger, more modern 
system to more adequately support the increasing use of this 
system by laboratory scientists. Work is already underway for 
the addition of high-speed CRT support by mid 1974. A new 
telephone switch specifically for data traffic is being installed 
by the communications office to remove this burden from the 



22 



N^laJtJ., v'oi.c:e swit.chboaiu» Efi-orts are also underway to have an 
uninterruptable power plant installed to insulate the computer 
system from laiiure in Lhe commercial power supply. 



23 



Serial No. 2.1 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title; Computer Representation and Manipulation of Macro 
Molecular Electron Density Surfaces 

Previous Serial Numbers None 

Principal Investigators R. J, Feldmann 

Other Investigators; G. H. Cohen 

Cooperating Units; LBM-NIAr4DD 

Man Years; 

Total: ,3 

Professional; .3 
Others; 

Project Description; 

Objectives ; 

The determination of the atomic coordinates of macro 
molecules is a lengthy process. The construction of a 
stick (Dryding) model of the atomic coordinates is 
accomplished by using the electron density map in the 
Richard box. After the model has been generated the 
atomic coordinates must be measured. It is normal 
thereafter to apply a computer program to the coordinates 
to check and refine them. The refined coordinates must 
then be worked back into the model to determine if there 
is still an agreement between the electron density map 
and the refined model. 

It is the global intention of this project to be able to 
start with the data for an electron density map and then 
to construct and refine the atomic coordinate model 
completely inside the computer. 



25 



Methods Employed: 

The macro molecular display program has been modified to 
include the functions for the generation of partial and 
complete sections of electron density contours. The 
electron density map for Gamma Chyotrypsin and a 
partially refined model have been brought together in the 
computer display. 

Major Findings: 

The electron density contours are drawn as a line segment 
approximation to a closed curve, A typical 2.0 - 2.5 
angstrom resolution map requires that three contour 
levels be drawn. A typical polypeptide context requires 
15 to 30 planes of electron density included in an eight 
angstrom sphere surrounding the central atom in the 
context. The display image generated by so many contours 
is very cluttered. It is very difficult to distinguish 
whether a contour is an inner contour on an upper plane 
or an outer contour on a lower plane (or visa-versa) . 
When the atomic model is added it is difficult to 
determine the position of atoms in relation to the 
electron density contours. It comes down to the fact 
that the representation of surfaces by a line drawing 
display does not produce an easily perceived image. 

The conclusion is the same as we came to in respect to 
the display of the shape of macro molecular active sites. 
To draw surfaces, a surface drawing display is required. 
In addition to generating an easily perceived electron 
density surface, the problem of model construction would 
be simplified. If each atom is represented by a regular 
polyhedron then the position and conformation of the 
atoms can be determined by maneuvering the polyhedra to 
produce a maximum intersection with the electron density 
surface. Human visual perception is more adept at 
determining the degree of intersection of two solids 
(surfaces) then at determining the centrality of a point 
in an amorphous volume. f 

Significance to Program of the Division: 

Several macro molecules are in the process of structure 
determination at the NIH, The use of the computer 
display should shorten the time and manpower required to 
solve a structure. 



26 



Propo^ea Coursa: 

Gtuclles ax-e under way to determine the feas-ibility of 
impienieixting a surface drawing display system. This 
scirfaca display system in addition to being used in the 
display of macro molecules could be used in the 
reconstruction of cellular entities from optical or 
electron microscopy sections, 

rtOivors axiC AA'ards; iMcne 

Piiblis?atjLcns: Nona 



27 



Serial No. 2.2 

1. Computer Centei Branch 

2, 

3. Bethesda 



PKS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 



Project Title: Comprehensive Communications Control 

Previous Serial Number: None 

Principal Investigator: G, Ricart _ 

Other Investigators: C. Henderson, K. Stevens 

Cooperating Units: None 

Kan Years; 

Total: 0.5 

Professional: 0.5 
Others : 



Project Description: 
Objectives: 



Computers are potentially most valuable in an on-line 
environment. The dynamic control necessary in such an 
environment calls for computer communications systems. 

Communications systems are the links between man, 
machine, and his computational system. Intimately 
involved are the human engineering and technology that 
provides realistic and controllable computer response to 
a human inquiry. 

The man/machine interface begins at the input/output 
device used. Interactive systems have "terminals" for 
users. Gradually, these tools have increased in 
performance so that, today, a computer can present a 
remote user with information as quickly as that user can 
read it' — and many times it will even be faster than that. 
This technology explosion has a serious impact on the 
interactive computer user. There is an extremely wide 
range of terminals currently available. Standards are 



28 



practically non-existant. Users are limited to those 
terminals which their interactive system will support. 
Due to the existing and unyielding nature of these 
present systems, the user is often held back by the 
forced choice of terminals. For example, the IBM 2741 
terminal y although in wide use throughout the campus, 
operates at only a fraction of the speed possible. Many 
people can read much faster than the 147 words~per-minute 
rate of such a device. New technology, (for example, the 
Diablo HyType), is available to replace the 2741 while 
still providing the hardcopy that makes the 2741 a 
natural choice. The IBM system will not support the 
HyType, and therefore, mantime is wasted in waiting for 
the old technology to function. 

One objective of this project is to bring a flexibility 
to the type of terminal that a user can select. This 
will allow the user to select the terminal best suited 
for his needs rather than having to use whatever his 
system used to support. 

Communications systems in use today also support fixed 
and rigid use of the terminals. Typically, only one 
computer service can be reached for a given type of 
terminal at one phone number. When that number is 
dialed, the user must observe a rigid format determined 
by fixed user software. 

Another objective of this project is to bring flexibility 
and human-kindness into the user of terminals. One phone 
number will suffice for all terminals (and if possible, 
services). The user will be able to specify how he wants 
to use his terminal device in communicating with the 
service. 

Locality is very important in determining modes of 
communications. Cost and effectiveness trade-offs reveal 
an entire spectrvun of reasonable communications modes. 
At present, rigid systems do not permit any variety in 
modes of operation. 

Another objective of this project is provision for 
expansion into new media as technology advances and 
cost-effectiveness show new directions. 

Some of these modes of operation will support 
computer-to-computer interfaces in networking operations 
that appear much like the ARPAnet. 

This project will pave the way for computer/computer 
networking. 



29 



Methods Employed; 

The terminal itself is the most basic form of man/machine 
interface. The investigators are maintaining a very 
close watch on developments in the field of terminals. 
This is being done to allow the project to sort out the 
clearly superior methods for recommendation for inclusion 
in the terminal network portion of the project. 

Other developments in the field of communication are also 
being watched closely for possible application. The 
result should be a state-of-the-art communications system 
that will provide the most effective human interface 
allowable under current technology. 

Work towards producing a comprehensive computer interface 
is underway on a PDP-11/20 computer interfaced to the 
PDP-10 timesharing system. This computer is equipped 
with a synchronous line interface which is used at the 
present time as a concentration point for several 
terminals located physically in Building 36 in the Lipkin 
Laboratory. This exercise in communications has produced 
an expertise for entering the more ambitious projects 
outlined here. 

Unfortunately, due to NIH red tape and supplier red tape, 
the hardware necessary to implement a terminal 
communications system is still in the distant future. 
Hardware for preliminary testing is installed and 
investigations and experiments based on this equipment 
are taking place. 

Significance to Program of the Division: 

Regardless of the rationale for the user of computers, if 
computers should be used, they should be used well. 
People should not be penalized for trying to take 
advantage of a computing system by being tightly bound in 
their choice of terminals, services, computer/computer 
interfaces, and their uses. This project is intended to 
attempt to remove as many stumbling blocks as the 
state-of-the-art will allow. 

Honors and Awards : None 

Publications: None 



30 



Serial No , 2 « 3 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

Jcly 1. 19 72 through June 30, 1973 

Project Title; Remote Computer Graphics 

Previous Serial Niambers None 

Principal Investigator: K. A. Stevens 

Other Investigators; R. Sproull 

Cooperating Units : None 

Mcui Years s 

Total? 1 

Professional: 1 
Others t 

Project Descriptioaj 

Objectives t 

The development of a generalized system for remote 
computer graphics. The generality of this system stems 
from its ability to interface to several user languages? 
handle various hardv;are devices? and lend itself readily 
to a 3.arge class of applications. 

Methods Employed: 

1, The Omnigraph system was designed and implemented to 
allovv' users to v/rite graphics programs in one of three 
languages (FORTRAN, SAIL, or LISP) and present graphic 
data to either a DEC 340, ARDS, or a Tectronics 4010 
display terminal. Plotting capabilities have been 
included for off-line as well as on-line plots. 

2. Continuing development of Omnigraph software. This 
effort includes modifying the existing code in response 
to user feedback as well as extending the overall 
capabilities of the system. 



31 



3. Encouraging the user community toward the development 
of remote graphics applications programs. 

4. Investigating state-of-the-art hardware and software 
which have potential applicability. 

Major Findings; 

1^ The basic software capabilities are adequate for most 
graphics applications encountered at: DCRT. 

2. Graphical interaction is important in most appli- 
cations ^ however ; hardware facilities for such are weak 
(limited to the tablet service provided with the DEC 
340), Currently, remote terminals merely provide 
keyboard interaction. 

3. In addition, remote graphics suffers from the low 
transmission rates provided by dial-up telephone lines. 
The display-generation time typically makes a graphical 
dialogue ponderous, and light-pen interaction unfeasible, 
if capabilities are not provided for local processing of 
basic graphics housekeeping (local display refreshing, 
interrupt handling, etc.). 

4. In applications that do not require interaction, ade- 
quate hardware capabilities can be provided by low-cost 
storage-tube CRTs such as the 4010. These devices 

are limited in resolution and picture quality, however. 

Significance to Program of the Division: 

The remote graphics effort at DCRT provides a valuable 
tool to the scientific community, for the scientist can 
manipulate graphical data through a variety of higher- 
level languages in a lov/-cost timesharing environment 
from a remote site (typically his lab) . 

Proposed Course: 

Continuing maintenance and support. Inclusion of GT40's 
into the Omnigraph system in order to provide improved 
remote graphics. The GT40 display consists of a refresh 
screen CRT coupled to a DEC PDP-11 minicomputer. This 
device allows local graphics processing, thus providing 
more efficient interaction. 

Improved telecommunication hardware within the host 
computer v;ill allow more remote graphics service. 

Honors and Awards: None 

Publications: None 

32 



serial No. 2.4 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title? PDP-10 Accounting System 

Previous Serial Number: None 

Principal Investigator: R. G. Dunham 

Other Investigators: G. Ricart 

Cooperating Units: None 

Man Years : 

Total : 1 
Professional: 1 
Others : 

Project Description: 

Objectives : 

The development of a computer system to automate 
fee-for-service accounting on a timesharing system. 

Methods Employed: 

This system exploits software on the PDP-10 timesharing 
system which automatically initiates programs. At system 
loadtime, a data gathering program is initiated which 
periodically records user activity throughout 
timesharing. A governing program then decides on the 
basis of the current date which account operations to 
schedule for the day (data reduction, backup, report 
writing, etc.), initiates the appropriate programs, 
instructs the operators to mount and dismount magnetic 
tapes/disk packs, and reports the results by a computer 
mail system to administrators of the accounting system. 



33 



Major Findings: 

Personal intervention has been significantly reduced. 
Sufficient communication regarding errors, income earned, 
and successful completion is maintained by means of the 
computer mail system. 

Significance to Program of the Division: 

This accounting system contributes to DCRT's experience 
in administrating timesharing services. 

Proposed Course: 

Maintaining and refining this system. 

Honors and Awards: None 

Publications: None 



34 



Serial No . 2.5 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July I. I'D 7 2 through June 30, 1973 



Project Title: Implementation of the Caiiibridge Crystal File 

Previous Serial Number: None 

Principal Investigators R. J. Feldmann, J, V, Silverton 

Other Investigators: S. R, Heller 

Cooperating Units: LC-NHLI 

Man Years ; 

Total: .2 

Professional: .2 
Ochers; 

Project Description: 

Objectives : 

The Cambridge (England) Crystal File contains all the 
bibliographic and structure information of all crystal 
studies published since 1960. The crystal file can be 
made to serve as a very useful reference tool for the 
crystallographers working at the NIH. 

Methods Employed; 

The molecular structure representation, searching, and 
display techniques previously developed were applied to 
the implementation of the crystal file. 

Major Findings: 

Very few programatic modifications had to be made in 
implementing the crystal file. 



35 



significance to Program of the Division: 

The implementation of the crystal file can be looked upon 
as another instance where techniques developed in the 
division could be applied to solve problems and provide 
tools for NIH scientists. 

Proposed Course : 

The crystal file will be updated periodically. New 
features will be added to the search- and display programs 
as the tool finds more use. 

Honors and Awards s None 

Publications: None 



36 



Serial No. 



Jl^A. 



1, Computer Center Branch 

2, 

3. Bethesda 



PHS - KIH 

Individual Project Report 

July 1, 1972 through June 30, 197 3 



Project Titles PDP-IO Systems 
Previous Serial Number: 2.11 
Principal Investigator: C. Henderson 
Other Investigators: G. Ricart, K, Stevens 
Cooperating Units : None 
Man Years : 

Total: 2 

Professional: 2 
Others : 

Project Description: 

Objectives; 

Development and expansion of techniques and methods to 
make the PDP-10 computer better suited to handle the 
computational needs of users for highly flexible, 
interactive systems. 

Methods Employed: 

1. Continuing development and maintenance of the PDP-10 
monitor supplied by Digital Equipment Corporation. This 
timesharing monitor has required several augmentations 
in order to meet NIH needs. Further, the manufacturer 
periodically updates this system, which requires that the 
new system be installed, together with NIH modifications. 

2. Continuing development and support of a library of 
systems programs to aid users. For example, the SOS text 
editor has been augmented to meet several new requirements, 



37 



3. Continuing and renewed support of interactive 
graphics capabilities using the PDP-10, The service 
programs for the 340 Display and RAND tablet have been 
redone in order to support more highly interactive 
systems with less load on the timesharing system. Only 
the portion of a user program which actually drives the 
display (the "display file") is retained in core; the 
balance may be swapped to secondary storage if other 
timesharing users require more immediate service. 

4. Adding communications support to allow remote 
computer and terminal tie-ins for symbiotic usage of the 
PDP-10. These additions have required procurement of 
hardware for high-speed communications. Software has 
been developed to allow high-speed synchronous 
communications via a PDP-11 acting as a communications 
controller/processor . 

5. Encouragement of users to develop specialized packages 
to suit a particular class of needs. Sometimes these 
programs gain wide usage, as have MLAB (A Modeling and 
Graphics Package), and R. Feldmann's Structure Search 
Package. 

Major Findings: 

1. It is quite feasible to attempt systems modifications 
with the DEC system. 

2. Ways ought to be investigated to better coordinate 
NIH software changes with software updates from DEC. 

3. The PDP-10 courses offered through the DCRT/CCB 
Training Section have been enthusiastically received and 
offer a wide variety of courses, ranging from learning to 
program to very sophisticated graphics and programming 
languages like APL. 

4. Having shifted from a laboratory to a fee-for-service 
environment, the PDP-10 has seen a great deal of growth, 
both in the number of users and their diversified 
applications and the machine utilization; with uptime 
expanding from eight to sixteen hours per day. As 
drastically as the machine usage has grown, so have the 
demands on the systems staff to keep the system current 
and to provide user support. Current staff is sufficient 
to only hold the line in the face of growing demands. 



38 



Significance tc Program of the Division; 

The PDP-IO has supported, various DCRT programs with 
timesharing services as well as unique programming 
languages and graphics support. In addition, various 
smaller laboratory computers are supported by DCRT on the 
PDP-10 for program assembly and computational support. 

Piuposeu Course j 

Continuing maintenance and support with expanding systems 
development as time and manpower permit j 

1. Implement existing plans for expansion of PDP-IO 
computing services in both software and hardware. The 
current KA'-IO processor is to be replaced by a dual 
KI-10 processor configuration with new high-speed 
memory and faster and more modern peripherals. hlso 
move up to DEC's most recent monitor release to 
incorporate new and better system facilities. 

2. Implement plans for a pseudo "third segment" 
feature with Omnigraph display routines. Implement 
MANTIS, a sharable and interactive FORTRAN debugger, 
along with the upgrading of remaining system 
Compilers/Utilities/Programs to reentrancy wherever 
possible. 

3. Improvexnent of communications by increasing 
asynchronous line capacity and by expanding synchronous 
communications development ^ both software and hardware. 
Implement a generalized communications scheme to handle 
a variety of terminals, including IBM 2741s, at various 
baud rates. 

4. Expand support to handle a larger and more 
diversified field of display devices from the currently 
supported DEC 360, Computek 400, and Tectronics 4010 
displays. This will include DEC's GT40 display and 
other devices as they become available and merit 
DCRT/CCB/PDP-10 support. 

Honors and Awards: None 

;?ublicatioris ; None 



39 



Serial No . 2.7 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Macro Molecular Display and Manipulation 

Previous Serial Number: 2.16 

Principal Investigator: R. J. Feldmann 

Other Investigators: J.S. Cohen, C.R.T. Bacon, S.R. Heller 

Cooperating Units: LCB-NIAI'IDD, LMB-NIAMDD 

Man Years: 

Total 0.7 
Professional: 0.7 
Other: 

Project Description: 

Objectives: 

The improvement of the operating characteristics of the 
macro molecular display system. 

Methods Employed: 

The display system written during the previous year was 
completely rewritten as a result of extensive use by the 
scientific community. The new program written in the 
programming language SAIL has the ability to expand the 
core size as a function of the number of atoms being 
displayed. Whereas previously it was impossible to 
handle molecules as large as the tetramer of Hemoglobin 
(4500 atoms), now for the few instants that user is 
operating with the tetramer, the program is occupying 
almost all the core resources of the PDP-10. As soon as 
the user selects a small subset of the tetramer for 
further observation, manipulation and display, the 
program contracts to a much smaller size. 



40 



In addition further improvements have been made to the 
display system in the areas of plotting and generation of 
molecular conformations for energy calculations. During 
this year eight films of macro molecular structures have 
been generated. These films have been used extensively 
as support for seminars in the NIH and at various 
meetings in this country and abroad. Filming techniques 
have developed to a point v/here any scientist can 
generate a ten minute film with about two man days of 
work . 

Significance to Program of the Division; 

The capabilities of the AGT-30 display terminal and the 
PDP-10 computer have been utilized to generate an 
operational macro molecular display system. In doing so 
we have brought to the NIH almost all the macro molecules 
which have had their crystal coordinates determined. 
This process of system generation, file collection and 
system use has brought the NIH to the state-of-the-art in 
molecular displays. 

Proposed Course: 

The AGT-30 display terminal which generates images by 
drawing lines is completely suited for drawing stick 
(Dryding) representations of macro molecules. Much of 
the interest in the structure of macro molecules is 
centered on the relationship of the structure of the 
active site(s) to the function of the molecule. For this 
work the shape of the active site conveys more infor- 
mation than the relationship of the atom centers. 

The line drawing AGT-30 display cannot adequately display 
the atomic surfaces. We have made a number of experi- 
ments about the representation of surfaces by connected 
line segments. The results are uniformly disappointing. 
Line segments which are in the back of a surface tend to 
confuse the perception of the shape of the front surface. 
The way to solve this problem is to use a display which 
draws surfaces rather than lines. One experiment has 
been made with an existing surface drawing display. This 
experiment indicates that the shape of a molecular 
surface can be easily perceived. The complexity of the 
shape of the surface is not an absolute hindrance to 
perception as the surface can be colored to convey 
recognition of atomic constituency. In addition, smooth 
shading can be used to create a sense of surface 
smoothness; and shadowing, caused by a light source, can 
be used to convey shape information. 



41 



studies are underv;ay to determine the feasibility of 
implementing a surface drawing display system to be used 
in the determination of the shape of active sites of 
macro molecules. 

Honors and Awards : None 

Publications: 

Feldmann, R.J.,- Heller, S.R.; Bacon, C.R.T. - An 
Interactive, Three-Dimensional Display, Manipulation and 
Plotting System for Biomedical Research, Journal of 
Chemical Documentation, Volume 12, No. 4, 1972, pp. 
234-237. 



42 



Serial No . 2.8 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: On-line File Processing System 

Previous Serial Number t 2,13 

Principal Investigator: Jon A. Halverson 

Other Investigators: flarvin Katz 

Cooperating Units: None 

Man Years ; 

Total: 2 

Professional: 2 
Others : 

Project Description: 

Objectives 3 

The primary objective is to establish an on-line file 
processing capability at the NIH v/hereby the user can 
maintain, retrieve and display information in a data 
base via a remote terminal. 

Methods Employed: 

The monitor being used is a modularly constructed program 
designed to provide the basic control system program for 
the installation of on-line information systems. It 
functions as an interface between user-written 
processing programs and the IBM system/360 operating 
system. 

Major Findings: 

The monitor tc provide the data base/data communications 
interface has been implemented. To evaluate its effec- 
tiveness, the activities of the NIH Clinical Center 
Admissions Department was chosen as the first 



43 



application. This subsystem is now operational. After a 
year-and-a-half now in operation the system has performed 
quite v/ell. Based upon experience to date the system 
should be considered as the vehicle with which to develop 
a Total Hospital Data Entry System. 

Significance to Program of the Division: 

The system can be considered as an initial step toward 
the development of on-line information systems for the 
NIH. For the first time investigators responsible for 
maintaining large data bases will have a powerful means 
of accessing critical information quickly. 

Proposed Course : 

Attention will be given to operating the system in an on- 
line, multi-user environment. Special emphasis will be 
given to applying the system to other areas in the NIH 
Clinical Center, in addition to Admissions, which can 
benefit from interactive processing. 

In particular, a generalized nursing unit system is 
currently being designed and implemented as a further 
extension of the Clinical Center on-line information 
system. 

An application has also been implemented in the 
Department of the Navy to further extend the on-line file 
handling capability of the monitor. Another application 
is being planned for NIH personnel. 

Honors and Awards : None 

Publications: None 



44 



Serial No. 2 . 9 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS ~ NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Titles Virtual Machine Facility 

Previous Serial Number: None 

Principal Investigator; Elliott Alterman 

Other Investigators s John Camp 

Cooperating Units: PAL Unit, Systems Team, Operations Section 

Man Years: 

Total: .8 
Professional: .6 
Other : . 2 

Project Description: 

Objectives : 

Using VM (the Virtual Machine facility) available for 
System 370 we will provide an environment wherein several 
copies/versions of the operating systems we run will be 
concurrently operational on the same processor. In this 
way individual changes to the system can be designed and 
debugged by several people at the same time v/ith no 
interference. As an added benefit we will gain experi- 
ence in the area of paging systems characteristics. 

Methods Employed: 

VM provides the vehicle for creating the environment in 
which multiple operating systems can function con- 
currently. By suitably defining the requirements of 
each "virtual machine" to VM, resources are 'virtually' 
available to that "virtual machine". For example, the 
"virtual machine" may think it has a real printer at 
address OOE when in fact that printer is simulated by VM. 



45 



Major Findings; 

To date some of the goals have been realized. We have 
tested changes to operating systems and debugged them 
without affecting the production environment. 

With some additional effort, we will shortly be able to 
run multiple operating systems simultaneously. 

Significance to Program of the Divisions 

Without a facility ^uch as this it is often necessary to 
test system changes on live systems with the resultant 
chances of system crashes. The "Virtual Machine" 
environment will allow system upgrade with minimal inpact 
on the production environment. 

Honors and Av/ards: None 

Publications: None 



46 



Serial No. 2.1Q 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title:: DATASTOR Program 

Previous Serial Number; 2.6 

Principal Investigator: James Ober thaler 

Other Investigators: William Jones 

Cooperating Units: None 

Man Months: 

Total: 1/2 

Professional: 1/2 
Others: 

Project Descriptions 

Objectives: 

1, To provide specifications for a general purpose 
process6r-to-processor communications facility. 

2. To implement, using these specifications, a set of 
computer programs which will permit computers remote from 
the main CCB facility to create, manipulate, and retrieve 
data files on the Computer Center's disk storage devices. 

Methods Employed: 

1. Communications links between the central and remote 
sites may utilize switched network or leased line 
facilities; both synchronous and asynchronous trans- 
mission techniques are supported. 

2. The programs which compose the DATASTOR system are 
coded in System/360 Assembler Language. 

3. The progrcims operate in the batch environment at 
present, as opposed to being a permanently on-line 
system, such as WYLBUR. 

47 



Major Findings: 

A Univac DDP 516 at the Dental Institute (Building 30) 
used the DATASTOR system for a brief period of time. No 
work is currently in progress. 

Significance to Program of the Division: 

This is an important step in the Computer Center's plan 
to create an environment in which the many small 
specialized laboratory computers can 'easily converse with 
the powerful central facility. 

Proposed Course : 

1. To complete the implementation of the system according 
to the specification developed. 

2. To develop a facility within the framework of the 
system for remote computers to submit batch jobs and 
receive their output. 

3. To institute permanently on-line service by the 
system, as soon as its utilization is sufficient to 
justify doing so. 

Honors and Awards : None 

Publications: None 



48 



Serial No. 2.11 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: HASP/Shared Spool 

Previous Serial Nxunber: 2.3 

Principal Investigator: William Jones 

Other Investigators: James Oberthaler 

Cooperating Units: None 

Man Years 

Total: 1.6 
Professional: 1.6 
Others: 

Project Description: 

Background : 

The HASP system is an extension to the principal control 
program employed on each of the Computer Center's IBM 
System/360 computers. HASP is responsible for reading 
all jobs into the multi CPU system, scheduling their 
execution on a priority basis, and handling all printed 
and punched output from the jobs. HASP provides the 
interface whereby conversational terminal systems enter 
programs into the batch stream and control their 
execution. 

Objectives: 

1. To tailor the standard HASP distributed by IBM to the 
specific needs of the NIH computing community. 

2. To integrate the NIH modifications to HASP into the 
new versions of HASP which are released periodically by 
IBM. 



49 



Methods Employed: 

HASP is a computer program written in IBM System/360 
Assembler Language. The standard version distributed by 
IBM consists of approximately 10,000 instructions to 
which NIH has nov7 added on the order of an additional 
10,000 instructions. 

Major Findings: 

A complete rewrite of HASP by the IBM implementation team 
necessitated an overhaul of the entire package of 
modifications which NIH has made to HASP. This new 
version of HASP v/as installed at NIH during May 1972. 

HASP-TSO interfaces were designed and implemented in 
January 197 3. The NIH modifications were integrated into 
a new HASP release in December 1972. Using an IBM 3705, 
7200 baud lines were installed for some high-speed remote 
terminals. The checkpointer process was rewritten to 
provide better protection against CPU failures. 

Significance to Program of the Division: 

The HASP system is of prime importance to the entire 
batch processing system at CCB, because HASP is signifi- 
cantly easier to modify than the Operating System, and 
is not sensitive to changes made in the Operating System. 
It contains many provisions that are unique to NIH. 

The multiple computer environment at the Computer Center 
allows significantly better overall service to the NIH 
computing community by providing both increased capa- 
bility and superior reliability to a single machine 
operation. Within this framework, however, the task of 
managing three computers rather than one presents 
formidable scheduling and logistical problems. Through 
the use of a shared input/output queue, many of the 
decisions currently being made manually can be automated, 
thus providing more immediate and accurate response to 
the variations in work load and resource availability. 

Proposed Course: 

To develop an adjunct to the HASP Shared Spool which will 
enable operators on a given system to issue commands to 
be executed on another system. This will support a 
considerable physical reorganization of the machine room. 



50 



Install a HASP-to-HASP remote facility that will allow 
distant CPUs to communicate over high-speed communication 
lines, adding further protection against disk hardware 
failures. 

Honors and Awards : None 

Publications: None 



51 



Serial No, 



2,12 



1. Computer Center Branch 

2 . 

3. Bethesda 



PHS - KIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: 360 Systems Development 

Previous Serial Number: 2.4 

Principal Investigator: Robert H. Brunelle 

Other Investigators: Staff of Systems Team 

1 
Cooperating Units: PAL Unit 

Man Years : 

Total: 4 
Professional: 4 
Others: 

Project Description: 

Objectives : 

To maximize the thruput and minimize turnaround time to 
all users of the central facility's 360 computers. To 
put computing power into every researcher's lab or office 
thru remote terminals and software systems. Develop the 
concept of the computer utility to its fullest. In 
general, provide the NIH user with the best Computer 
Center and service he can get thru software development 
and hardware expansion. 

Methods Employed: 

Judicious selection and tailoring of software systems to 
the NIH environment such as TSO and CICS. Acquisition 
hardware that compliments the software to provide the 
researcher with the variety of tools he needs. Constant 
attention is given to the overall system software and 
hardware needs so that as new systems and hardware become 
available they are offered to our users when needed. 
Increasing the capacity of the system to keep ahead of 
the needs of the NIH users. 



52 



Significance to Program of the Division: 

Constant attention to all facets of computing provides 
the NIH with an up-to-date computer facility catering to 
the needs of all NIH's researchers. Computer power in 
the lab brings a powerful research tool closer to the 
projects, thus making it easier to use, and more likely to 
become an integral part of, the research program. Close 
attention to daily running permits us to tune the total 
system, maximizing throughput while minimizing turnaround 
time, thus remaining responsive at all times to the 
constantly growing workload. New hardware and software 
are offered to fill the varied needs of the community. 
All of these efforts combine to form a foirward looking 
Computer Center sensitive to the needs of the NIH 
research environment. 

Proposed Course: 

To monitor the NIH environment, continuing to mold the 
computer facility to best meet the needs of NIH. 

Honors and Awards : None 

Publications : None 



53 



Serial No._ 2.13 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: WYLBUR Text-editing System 

Previous Serial Number: 2.5 

Principal Investigator: James Oberthaler 

Other Investigators: Leslie Harden 

Cooperating Units: None 

Man Years: 

Total: 0.6 
Professional: 0.6 
Others: 

Project Description: 

Objectives: 

1. To provide Computer Center users with a convenient 
and comprehensive tool to aid in the creation and 
development of computer programs and data files. 

2. To make it significantly easier to create and edit 
text materials such as letters, reports, etc. 

3. To reduce the necessity for users to leave their 
laboratories or offices in order to use the computer, 

4. To provide a convenient method for ascertaining the 
status of the computing system as a whole and locat- 
ing particular jobs as they are being processed. 

Methods Employed : 

WYLBUR is an on-line text-editing and remote job entry 
system oriented to low-speed character mode terminals 
such as the IBM 2741, 1050, and 2260 and Models 33, 35 
and 37 teletype machines. It was originally designed 
and implemented at the Stanford University Computation 

54 



Center. iJYLBUR provides the user, in his lab or office, 
the facility to create and edit source programs in real 
time, submit them for compilation and execution by the 
standard job stream processor, and retrieve the results 
of execution at his terminal. In addition, the user may 
create, store and modify text material, such as letters, 
reports, books, etc. ^'JYLBUR is constantly being modified 
to improve performance and make new facilities available 
to the users. 

WYLBUR consists of two computer programs written in IBM 
System/360 Assembler Language which reside permanently in 
two separate regions under OS/36 0. MILTEN, the tele- 
communications monitor, controls all input and output for 
the low-speed terminals and handles communication with 
the system operator. MILTEN can communicate with one or 
more subsystems operating in other regions under OS/36 0. 
All subsystems use the same pool of low-speed lines, thus 
reducing the cost and providing more flexibility for the 
user. WYLBUR, the text editor, is a subsystem of MILTEN. 
It performs all of the actual manipulation of text and 
the remote job entry to HASP. V/YLBUR calls on MILTEN in 
a device independent manner for all terminal I/O requests. 
The permanent text files are stored on 2 314 disks which 
are shared among all CPU's in the system. This permits 
batch jobs to access WYLBUR data sets. The text is stored 
in a compressed format in order to save space. The work- 
ing files are stored on 2301 drum (with disk as backup) 
and are moved in and out of core under a demand paging 
algorithm. 

Major Findings; 

WYLBUR has been extensively modified in the last year in 
order to increase its usefulness to users and to increase 
its capacity. 

lo The number of telephone lines into WYLBUR has been 
increased from 111 to 184 in order to keep pace with the 
constantly increasing demand. The maximum number of 
simultaneous users observed to date is 178. 

2, The number of online file disks has been increased 
from 36 to 43 in order to satisfy the increasing demand 
for storage space. 

3. WYLBUR has been modified to support Rotational 
Position Sensing (RPS) for its paging file. This 
modification greatly improved system responsiveness. 



55 



4. Accumulated charges for the session are displayed 
when the user signs off. 

5. A version of WYLBUR has been developed for use as 
an editor under the Time Sharing Option (TSO) system 
now supported at the Computer Center. 

In addition to the effort described above, there is an 
extensive program to educate the users about WYLBUR. 
Many sessions of the WYLBUR for Programmers and the 
WYLBUR for Secretaries and Administrative Personnel 
courses have been taught, and a WYLBUR Reference Card 
was published. 

Significance to Program of the Division: 

WYLBUR represents a significant move towards lowering 
the amount of non-programming overhead involved in the 
development of programs. With the job output available 
at the user's terminal, effective turnaround time is 
substantially lowered. In addition, the added ease of 
using a typewriter-like terminal instead of a keypunch 
contributes to more rapid development and debugging of 
programs. WYLBUR has also gained wide acceptance as a 
tool to aid in the preparation of memos, reports, and 
other documents. In addition, WYLBUR has been used for 
data collection and inquiry applications. 

Proposed Course: 

The Division is in the process of procuring a new version 
of WYLBUR, upward compatible with the installed WYLBUR, 
which will offer many new editing facilities to the 
WYLBUR user while at the same time enhance its efficiency 
and responsiveness under the heavy loads currently 
experienced at CCB. 

Honors and Awards : None 

Publications: None 



56 



Serial No. 2.14 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 197 3 

Project Title: CPS - Conversational Prograinming System 

Previous Serial Number: 2.10 

Principal Investigator: John Camp 

Other Investigators: Trudy Kenny 

Cooperating Units: None 

Man Years : 

Total: 1-1/2 
Professional: 1-1/2 
Others: 

Project Description: 

Objectives : 

CPS continues to provide much of the power of the central 
computer facility to the researcher in his lab or office. 
By using CPS the researcher has available the computing 
capability of the central 36 0/370 computers as well as 
the ability to access data bases stored on-line at the 
central site. 

Methods Employed: 

The CPS system allows multiple users to simultaneously 
write, debug, and execute 370 computer programs conver- 
sationally. Programs are written in a subset of the PL/1 
language from a 2741 typewriter terminal or teletype 
located in or near the user's office. The terminals are 
connected with the NIH central 36 0/370 system over 
telephone lines on a dial-up basis. In addition, CPS 
supports a dialect of the BASIC language and provides a 
remote job entry (RJE) facility through which jobs can be 
submitted to be run in the background batch environment. 



57 



Major Findings: 

Continuing enhancements to CPS place increasing power at 
the user's disposal. During the past year maintenance 
support has increased the reliability of the system and 
added additional power to the language. Further changes 
to the statement executor provided a clearer definition 
of how ranges of statements are executed. Improved error 
messages give the user a better explanation of what 
problem has occurred. 

NIH modifications to this IBM distributed program have 
enabled it to operate effectively as an integral part of 
the Computer Center. These modifications provide inter- 
faces for accounting, remote job entry, and data set 
naming. 

CPS uses a special shared catalog for its data set 
references so that batch jobs creating data sets for CPS 
can run on any machine. Since this is the same catalog 
used by TSO both systems can access the same data sets. 
A change to the terminal handler eliminated situation 
which caused the user's phone line to drop during login. 
Additional utility checking has improved progress 
security by insuring that only the owner can modify or 
erase a program. 

During the past year there has been continuing growth in 
the number of users and in the use of CPS. To meet 
increasing demands for access to CPS the number of ports 
for both 2741 and teletype phone lines was increased by 
37%. The average number of concurrent users has increased 
approximately 50% during the past year. 

Significance to Program of the Division: 

CPS continues to provide computing capacity in the lab or 
office allowing the researcher to utilize the central 
computer facility without leaving his normal working 
environment. Thus the computer is more readily available 
and much time is saved by not making the user come to the 
central site. By providing a truly interactive conver- 
sational environment CPS gives the user capabilities 
which could never be achieved in a strict batch mode of 
operation. 

Proposed Course: 

Intended support includes continuing to provide mainten- 
ance for CPS. New versions and updates will be installed 
as they become available and are needed to improve the 
reliability and performance of the system. Local plans 

58 



include steps to further integrate CPS into the operation 
of the central facility, improvement in user interaction 
with CPS, and improved performance. 

With the increasing load on the system increasing demands 
are placed on system resources, particularly memory in 
order to solve this problem a major effort is in prooress 
to convert to a version of CPS which runs as a processor 
under TSO. In addition to providing a better base for 
CPS the new system will provide new user facilities. 

In order to prawi.de additional terminal support an 
investigation is in progress to determine the feasibility 
of offering high-speed CRT terminals (at 1200 baud) for 
CPS use. 

Honors and Awards: None 

Publications : None 



59 



Serial No. 2.15 .__ 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Optical Document Processing System 

Previous Serial Number: 2.14 

Principal Investigator: Elliott Alterman 

Other Investigators: John Camp 

i 
Cooperating Units: None 

Man Years: 

Total: 0.1 
Professional: 0.1 
Other: 

Project Description: 

Objectives: 

NIH processes vast volumes of data in many forms and 
formats. The effort necessary to analyze and process the 
data currently requires significant time and effort to 
convert this data to machine readable form. We have 
eliminated much of this intermediate data conversion by 
using the documents themselves as input to the central 
facility computers. 

Methods Employed: 

Internal processing routines have been developed to pro- 
cess documents read by the optical character recognition 
equipment (IBM 12 88 Optical Page Reader) . Each document 
is described uniquely in terms of a set of parameters. 
This information is maintained in an on-line data set 
accessible to the processing program. The flexibility of 
this design is displayed in the elmination of the 
requirement for writing new processing programs for each 
new document we wish to process. Changes to document 
design, processing options, or output formatting can be 
done in a matter of minutes. 

60 



Major Findings: 

This system permits rapid and efficient collection of 
data for processing by a computer. Significant cost 
reductions have been achieved in the year that the system 
has been functioning. Coupled with the great increase in 
data availability this has yielded a high performance, 
low cost data collection tool that permits greater 
utilization of available data. 

Significance to Program of the Division: 

The Division of Computer Research and Technology has a 
major goal of making access to the central computer 
facility easy, simple and cost effective. The Optical 
Page Reader System provides an avenue for that access in 
a manner requiring minimal change to external operating 
procedures. 

Project Status: 

System has been fully operational for 1^ years. Current 
status of project is one of maintenance only. 

Honors and Awards: None 

Publications: None 



61 



Serial No. 2.16 



1. Computer Center Branch 

A. . 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 



Project Title: Computer Output Microfilm 
Previous Serial Number: 2.18 
Principal Investigator: William Speary 
Other Investigators: W. Jones, R. Baxter 
Cooperating Units: None 
Man Years : 



Total: 


1/8 


Professional: 


1/8 


Others: 






Project Description: 

Objectives : 

To provide the NIH community a mechanism to obtain high 
volume output in a manner by which the output can be 
easier stored, manipulated and processed at a considerable 
cost and manpower saving. 

Me t hod s Emp 1 oy e d : 

A Datagraphix Micromation Recorder Model 4460 has been 
acquired by DCRT/CCB. General-purpose software routines 
have been written to allow the NIH community to obtain 
output which now is produced on paper to be produced on 
16mm microfilm or 105mm microfiche. Microfilm pro- 
gramming courses are also being offered to the NIH 
community. 

Significance to Program of the Division: 

The new COM facility gives NIH computer users the ability 
to automatically produce their computer generated output 
on either microfiche output in upper and lower case at 



62 



12,000 lines per minute, over 14 times faster than the 
traditional compiiter impact printer. The COM unit 
installed at the Center is a Stromberg Datagraphix 
SD-4460 Business Graphics Recorder attached directly to 
the central 360/370 systems. The 360/370-4460 
capabilities include not only high-speed "printing", but 
also automatic forms flashing, high resolution graphics, 
and universal film formats of 16mm and 105mm. The • 
primary advantages of COM over conventional impact 
printers are throughput speed, rapid information 
retrieval, low cost, unlimited number of copies, ease of 
handling and storage, and ease of distribution. 
Throughput increases from 900 lines per minute on the 
printer to 12,000 lines per minute using COM output. 
This can eliminate a considerable amout of delay in 
printing large volume reports. COM based files can store 
a large volume of information in a small space, and 
access to specific data via microform readers can take 
place in just seconds. The NIH software provides 
automatic titling, sequence numbering and an index for 
data output to microfiche. Depending on the voliime of 
printing, COM based systems can significantly reduce 
report producing and handling costs, thereby reducing the 
total data processing cost. An unlimited number of 
copies can be produced, each with the same quality as the 
original. This is in contrast to the maximum of four 
copies which can be "legibly" produced from the impact 
printer. 

Honors and Awards; None 

Publications : None 



63 



Serial No. 2.17 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Installation of Time Sharing Option (TSO) of 
OS/360 MVT 

Previous Serial Number: 2.19 

Principal Investigator: William Speary 

Other Investigators: A.C. Curtis, M. Nash 

Cooperating Units: None 

Man Years : 

Total: 1/2 

Professional: 1/2 
Others: 

Project Description: 

Objectives: 

To provide a Time Sharing System based on central 
computing facility for the NIH community to enable them 
to implement data processing programs in a shorter period 
of time. 

Methods Employed: 

OS/360 MVT-TSO has been installed in the NIH central 
computing facility. It was announced to users on 
November 10, 1972. TSO programming courses are also 
being offered to the NIH community. 

Significance to Program of the Division: 

Laboratories will be able to have, via terminals, 
computing facilities which have heretofore been available 
only through the 'batch' computer service. TSO also 
allowed the laboratories to program/debug/implement 
programs via terminals which are unique to the individual 
laboratories. 

64 



Proposed Course: 

CCB is in the process of modifying/enhancing TSO in order 
to operate in the NIH environment. TSO will allow 
users to use the central facility in a manner which has 
never before been possible — to construct, test, and 
execute programs interactively from a terminal in the 
language of the user's choice. TSO fills the gap between 
the fast-turnaround batch service and the powerful 
editing capabilities of WYLBUR and offers facilities 
differing dramatically from those offered by other 
currently available interactive systems such as CPS. 
Most importantly, TSO creates a timesharing environment 
for interactive and terminal oriented applications, and 
will allow users to develop conversational programs and 
systems which suit their individual needs. 

Honors and Awards : None 

Publications: None 



65 



Serial No. 2.18 

1 . Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1972 through June 30, 1973 

Project Title: INTERFACE (technical communications) 

Previous Serial Number: 2.8 

Principal Investigator: Joseph D. Naughton 

Other Investigators: Staff of Systems Team and PAL Unit 

i 
Cooperating Units: None 

Man Years: 

Total: 2 

Professional: 1 
Others: 1 

Project Description: 

Objectives : 

1. The primary purpose of this effort is to give the 
Computer Center a regular means of conveying, (1) highly 
current technical information to the widely dispersed 
computer programmers and systems analysts at NIH, and (2) 
current Center accomplishments, plans, policies, course 
offerings and other information of interest to personnel 
and general managers at NIH. 

2. A secondary purpose is to foster communication and 
collaboration among computer users and between them and 
the Center by providing a forum for: (1) announcing key 
new projects, accomplishments, personnel or organ- 
izations; and (2) airing viewpoints or suggested 
approaches to computing problems. 

Methods Employed : 

INTERFACE is published every four to six weeks, and 
distributed to all scientific and administrative 
personnel who have expressed a desire to be kept up-to- 
date on activities and services of the NIH Computer Center. 

66 



Significance to Program of the Division: 

INTERFACE complements the Users Guide and other technical 
manuals by highlighting items and directing readers to 
the other publications for details. INTERFACE has given 
the users of the Computer Center a single reference point 
for all communications concerning the use of compu- 
tational services at NIH. It has done an excellent job 
of keeping the users informed of new services, major 
systems changes and all facets of computing at NIH. 

Proposed Course : ' 

INTERFACE will continue to inform the Computer Center 
users of all pertinent computing highlights on a regular 
basis. However, due to manpower limitations the 
frequency of publications and amount of information 
included will be reduced somewhat this fiscal year. 

Honors and Awards: None 

Publications: None 



67 



Serial No. 2.19 

1. Computer Center Branch 
2 

3. Bethesda 



PHS - NIH 
Individual Project Report 
July 1, 1972 through June 30, 1973 

Project Title: Computer Training Program 

Previous Serial Number: 2.12 

Principal Investigator: Rita G. Minker 

Other Investigators: Patricia L. Logan 

Cooperating Units: All DCRT Laboratories and Branches 

Man Years: 

Total: 6.7 5 

Professional: 6 
Others: 0.75 

Project Description: 

Objectives: 

1. To provide the NIH staff v;ith an understanding of and 
ability to use modern computer technology. 

2. To facilitate the formulation and analysis of the 
problems to be solved so that the technology can be 
usefully applied. 

Methods Employed: 

1. Training courses are tailored to the specific 
hardware and software available at NIH. 

2. Seminars focus on both the underlying disciplines 
(e.g., applied mathematics, computer-related engineering, 
information sciences) and on the specific areas of 
potential application. 

Significance to Program of the Division: 

Training courses enable investigators to use DCRT's 
powerful equipment. Effective use of advanced techniques 

68 



and advanced equipment is obtained by further training 
and directed discussion. 

During the past 12 months, over 2000 registrations were 
processed for the approximately 80 short courses offered. 

Results from the Computer Training Survey Questionnaire 
conducted in the fall of 197 2 indicates that the training 
effort has been successful. Responses of 833 students 
showed a very high positive correlation of DCRT computer 
training with job efficiency and effectiveness. This is 
advantageous to DCRT from two viewpoints: 

(1) It is valuable for the Computer Center Branch 
because it contributes to efficient use of the 
computer system. 

(2) It is valuable for every B/I/D at NIH because it 
contributes to effective and efficient use of people 
and money. 

Questionnaire responses from 304 supervisors indicated 
that they were aware of the need of computer training for 
their employees. Most supervisors who responded to the 
survey found that the DCRT training program helped their 
employees do their jobs more effectively. 

Proposed Course : 

Fall 1973 and Spring 1974 semesters of training classes 
and seminars are planned. 

Honors and Awards: None 

Publications: 

1. Computer Center Branch: Computer Training 
Courses — Fall Term 1972. U. S. Dept. of Health, 
Education and Welfare, Public Health Service, National 
Institutes of Health, Division of Computer Research and 
Technology, Computer Center Branch. 

2. Computer Center Branch: Computer Training Courses 
and Seminars — Spring Term 197 3. U. S. Dept. of Health, 
Education and Welfare, Public Health Service, National 
Institutes of Health, Division of Computer Research and 
Technology, Computer Center Branch. 

3. Computer Center Branch: "Introduction to CPS Files - 
A Self Study Guide" 



69 



Serial No._ 2»20 

1. Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: User Support and Communications 

Previous Serial Number: 2.1 

Principal Investigator: Frances E, Halverson 

Other Investigators: Staff of PAL Unit 

Cooperating Units: Systems Team, Operations Section, Training Unit, 

Technical Information Office 

Man Years : 

Total: 8 

Professional: 8 
Others: 

Project Description: 

Objectives: 

To provide the users of the Computer Center with the 
personal assistance necessary to make effective use of 
the Center's facilities. To provide users assistance in 
resolving problems encountered while using the systems 
(hardware and software) maintained by the Center. 

Methods Employed: 

The PAL (Programmer Assistance and Liaison) Unit was 
established to perform this function. Thru the Bugs, 
Diagnostics and Hints Section of the Computer Center's 
technical report, INTERFACE, and the Users Guide , the PAL 
Unit keeps the user community abreast of the latest word 
in problems, system changes and programming hints. They 
also assist the Systems Team in the design, selection 
installation, testing and modification of all system 
software provided by the Computer Center. Another method 
of written communication is the Programmer Trouble Report 
(PTR) . The Unit processed 1900 PTR's in 1972. This form 

I 
70 . 



allows the user to make complaints , suggestions or 
comments^ requests or to describe a problem he needs 
help with. The information from this form helps PAL and 
CCB to formulate future policies, plan systems changes 
and inform IBM or the Systems Team of software/hardware 
problems. Six of the nine staff members teach in 
the CCB educational program on a regular twice/year basis. 

Much of the time that is not spent working directly on 
users' programs is devoted to maintenance of the current 
system and on-line disk packs and to applying eind testing 
vendor-supplied fixes to the system. They are also 
responsible for testing and documenting nev? releases of 
the operating system. This involves developing and 
running the NIH test job stream against each new release 
to test its integrity. The test job stream, which is 
made up of jobs submitted by users, is intended to be 
representative of the work done at the NIH Computer 
Center. By evaluating the results of these runs, the PAL 
Unit is able to prepare users in advance for any 
necessary changes and to accomplish the transition from 
release to release with a minimum of inconvenience to the 
users. They evaluate new terminal systems and software 
packages being considered for implementation. 

Significance to Program of the Division: 

The computer user at NIH has at his disposal a group of 
competent professional programmers that are able to 
assist him with his problems in running programs or using 
the computer facility. The PAL Unit notes all trouble 
areas and, through INTERFACE and the PTR form, 
communicates common problem areas to all users. In all 
facets of computer use the user has someone that can give 
him the answer or guarantee to find the answer for him. 

Honors and Awards ; None 

Publications? None 



71 



Serial No, 2.21 

1, Computer Center Branch 

2. 

3. Bethesda 



PHS - NIH 

Individual Project Report 

July 1, 1972 through June 30, '1973 



Project Title: Computer Center Usex's Guide 

Previous Serial Number: 2.2 

Principal Investigator: Frances Halverson 

Other Investigators: Leslie Harden, PAL Unit and Systeras Team 

Cooperating Units: None 

Man Years: 

Total: 1 
Professional: 1 
Others: 

Project Description: 

Objectives : 

To provide the users of the central computer facility with 
a guide to the services, standards and use of the Computer 
Center. 

Methods Employed: 

A 400-page Users Guide was published and distributed in 
July 1972 to all users of the Computer Center, to 
Institute and Division Administrative or Executive 
Officers, to contract companies required to use the 
Computer Center and to other organizations and indi- 
viduals having a logical need for it. Updates to the 
Users Guide are published and distributed as necessary to 
keep it current. A major update was sent out in December 
1972. 



72 



Significance to Program of the Divisi 



on: 



For the first time all information pertaining to computing 
was brought together in a single reference document. The 
Users Guide contains a description of all facilities and 
services and how to use them. Programming standards, 
languages supported, JCL summary and other facets of com- 
puting are all given in detail. it is a complete guide 
to computing for the computer user. 



Honors and Awards: None 
Publications: None 



73 



July 1, 1972 through June 30, 1973 

PUBLIC HEALTH SERVICE - NATIONAL INSTITUTES OF HEALIH 
DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 

Sumnary of Branch Activities 1. DCRT - 3 

2. LABORATORY OF APPLIED STUDIES 3. Eugene K. Harris 

Chief 

The Laboratory of Applied Studies engages primarily in cooperative studies with 
NIH scientists, applying mathematics, statistics and con^juting science to bio- 
medical problems. In addition, the Laboratory imdertakes mathematical and 
statistical research with applications to biology. Some of the highlights of 
LAS studies of particular interest to clinical research and medical care are 
noted below. 

During this reporting year, the Medical Applications Section of the Laboratory 
has completed the first phase of its evaluation of currently available conputer 
programs for the interpretation of electrocardiograms. One of these programs 
has been selected for routine use at NIH in cooperation with the Cardiology 
Branch, NHLI. To date, coiiparison of programs has relied on agreement with 
independent assessments of ECG's by an NIH board of cardiologists. Current 
plans call for cooperative development of clinical documentation of specific 
diagnoses (ventricular hypertrophy and myocardial infarct) and further evalua- 
tion of coii5)uter algorithms to diagnose these diseases. In addition, conputer 
storage and retrieval facilities are needed to permit physicians to call up 
earlier ECG tracings and diagnoses on the same patient for review in conjunc- 
tion with examination of current data. 

Installation of a small corputer in the Nuclear Medicine Department, C. C, 
dedicated to the capture and analysis of data from the Gamma camera, has been 
conpleted. During the past year, this confuting system has been applied to 
studies of regional blood flow and dynamic mapping of renal uptake and excre- 
tion. Continued use in such research studies and applications to the diagnosis 
of brain lesions and disturbances of liver and kidney function are planned for 
the coming year. 

These studies of conputer applications to ECG analysis and medical use of radio- 
isotopes represent a growing investment by this Laboratory in the use of 
conputers to support and inprove non- invasive diagnostic procedures in medicine. 

During this year, the Applied Mathematics Section of the Laboratory perfected 
a general nonlinear, time -dependent model reported earlier to describe the 
diffusion of oxygen from capillaries to tissues. The model is now particularly 
useful for its ability to predict changes in oxygen concentration and distri- 
bution in tissue cells following disturbance to the normal oxygen-hemoglobin 
dissociation function under stress conditions. The numerical techniques and 
conputer programs developed for this oxygen diffusion problem are also expected 
to find application to other time -dependent diffusion processes in physiology. 



75 



statistical studies of nonnal variations in blood chemistries, which have been 
pursued during the past several years in this Laboratory, were largely concluded 
this reporting year. Extending previously reported studies, con5)uter-based 
methods have been published for the efficient analysis of "normal -value" 
distributions. During this year, investigation was begun on the application 
of time series analysis to the evaluation of short-term effects of drug therapy, 
studied through intensive time -dependent clinical trials. Such studies, as 
distinct from the long-term epidemdological type of follow-up trial, are 
frequently undertaken without paying adequate attention - in either the design 
or analysis stage - to the critical inq^ortance of correlation between successive 
observations. Continued applications of statistical' theory to time -dependent 
studies in hospital patients are planned for the coming year. 

During the past year. Laboratory staff conducted courses for NIH scientists in 
graph theory, use of the Calcomp plotters and use of the Modelaide programs for 
fitting mathematical models to data. 

Dr. James J. Bailey was appointed Head of the newly- formed Medical Applications 
Section. On the recommendation of the Director, DO^T, Dr. Harris' nomination 
for a research assignment abroad (London) has been approved to start early in 
FY 1974. 



76 



LAS reports published or "in press" during FY 73 

1. Ashbrook, J. D., Specter, A. A., and Fletcher, J. E.: Medium chain fatty 
acid binding to human plasma albumin. Journal o£ Biological Chemistrv 
Vol. 247, No. 21, Nov. 10, 1972. ^' 

2. Bailey, J. J., Hirshfeld, J. W. , Jr., Grauer, L. E. , Karsh, R. B. , and 
Itscoitz, S. B. : Unreliability of conputer programs in interpreting 
electrocardiograms. Circulation , Vol. 46: II-8, Oct. 1972. (Abstract) 

3. Bailey, J. J., Horton, M. R. , Pottala, E. W. : Sensitivity of ECG computer 
diagnostic programs to digitizing techniques. Proceedings of the 2nd 
National Conference of the Society for Con^Duter Medicine, Nov., 1972. 

4. DeMets, D. L. and Harris, E. K. : Transformation of observed distributions 
to Gaussian form. DCRT Tech. Rep. No. 8 , July, 1972. 

5. Fletcher, J. E. : On the effects of time -dependent flow in capillaries 
supplying substrates to living tissues. SIAM Review (Chronicle) Vol 15 
No. 1, January, 1973. ' ' 

6. Fletcher, J. E., Ashbrook, J. D., and Spector, A. A.: Conputer analysis 
of drug-protein binding data. Annals of the Ne w York Academy of Sciences 
1973. (in press) ~~~~ ^ ' 

7. Fletcher, J. E.: A mathematical model of the unsteady transport of oxygen 
to tissues m the microcirculation. Advances in Medicine and Biology 
Proceedings of the Third International Syirposium on Oxygen Transport to 
Tissue. Plenum Press, (in press) 

8. Gelfand M. J. and Clancy, D. L. : Comparison of indices of left ventricular 
(LV) performance in man. Circulation , Vol. 46(4): 11-55 Oct 1972 
(Abstract) ~ ' ' ' 

9. Harris, E. K and DeMets, D. L.: Estijnation of normal ranges and cumulative 
proportions by transfonning observed distributions to Gaussian form 
Clinical Chemistry , Vol. 18, 605, 1972. 

3. Harris, E. K. : Estimating correlations from single-sample distributions of 
measurements. Clinical Chemistry , Vol. 19, 81, 1973. 

L. Hutchinson, G.: The representation of lattices by modules. Bull. Amer 
Math. Soc , Vol. 79, 172-176, 1973. _uix^_™ner^ 



Hutchinson, G. : Recursively unsolvable word problems of modular lattices 
and diagram- chasing. J. of Algebra , (in press) 

Hutchinson, G.: On classes of lattices representable by modules 
Proceedings of the Houston Conference on Lattice Theory, (in press) 



77 



14. Larson, S. M. , Bailey, J. J., Schall, G. L,, Griffith, J, M. , and 
Johnston, G. S.: Radioisotope teclmique for measuring regional organ 
blood flow. International Journal of Applied Radiation and Isotopes, 
Vol. 23: 390-392, August, 1972. — 

15. Larson, S. M. , Millar, R. C, Chalmers, T. C, Kramer, R. T. , Bailey, J. J., 
and Johnston, G. S.: Quantitation of hepatic blood flow by xenon clearance: 
a primate model. In The Liver , Proceedings of the First GSTAAD, Sept., 
1972, G. Paumgartner and R. Preisig, editors. S. Karger Publishing Co., 
Basle, Switzerland. 

16. Mortimer, J. A.: Temporal- sequence of cerebellar Purkinje and nuclear 
activity in relation to the acoustic startle response. Brain Research , 
Vol. 50, 457, 1973. 

17. Mortimer, J. A.: A conputer model of mammalian cerebellar cortex. 
Conqjuters in Biology and Medicine , (in press) 

18. Pottala, E, W. , Colbum, T. R. , and Humphrey, D. R. : A dendritic compart- 
ment model neuron. IEEE Transactions Biomedical Engineering , Vol. BNffi-20, 
No. 2, p. 132, March, 1973. 

19. Pottala, E. W. : Conputer control of a conpartmental model neuron. Medical 
Instrumentation , Vol. 7, No. 1, 74, January, 1973. (Abstract) 

20. Pottala, E. W. : A hybrid neural simulation system. Proceedings of the 
1973 Summer Computer Simulation Conference , July, 1973. 

21. Pottala, E. W. , Justiss, R. G., Horton, M. R. , and Bruce, M. C: Multi- 
purpose conputer system for biological research. Proceedings of the 25th 
American Conference for Engineering in Medicine and Biology , p. 103, 1972. 

22. Pottala, E. W. : Computer controlled real time spectral analysis as a tool 
for processing physiological data. Medical Instrumentation , Vol. 7(1): 
74, Jan., 1973. (Abstract) 

23. Sharma, B., Pottala, E. W. , Zajac, F. E. III.: Analysis of the walking 
reflex via a corapartmental model neuron. Proceedings of the 25th American 
Conference for Engineering in Medicine and Biology , p. 67, 1972. 

24. Spector, A. A., Santos, E. C, Ashbrook, J. D., and Fletcher, J. E. : 
Influence of free fatty acid concentration on drug binding to plasma 
albumin. Annals of the New York Academy of Sciences , 1973. (in press) 

25. *Spector, A. A., Ashbrook, J. D., Santos, E. C, and Fletcher, J. E.: 
Quantitative analysis of free fatty acid uptake by mammalian cells: 
Laurie acid and human erythrocytes. J. of Lipid Research , Vol. 13, 
No. 4, July, 1972. 

*Listed as "in press" in FY 72 report. 



78 



Serial No. DCRT 3.1 

1. Laboratory o£ Applied Studies 

2. Biomathematics and Statistics 
Section 

3. Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Biormthematics and Statistics 

Previous Serial >&anber: Same 

Principal Investigator: J. E. Mosimann 

Assisted by: G. Hutchinson, G. Atta 

C. Clark (Project Stride) 

Co -Investigators: C. Merrill, NMI 

A. Coul ombre, NEI 

P. Martin, Univ. of Arizona 

J. Prewitt, D^IB, DCRT 

Man Years: Professional - 3.0 
Other - 0.2 

Project Description: 

Ob j ectives : 

To conduct mathematical statistical research in areas of biological 
in^Dortance, especially in multivariate problons, and to develop 
related con^juting methodolog)'- of use to biostatisticians and other 
health scientists. To investigate the applicability of mathematical 
theory, particularly finite mathematics including graph theory to 
varioiis biomedical problems. 

Progress during FY 73: 

Theoretical work continued on the development of multivariate mathe- 
matical methods for biological data vdiich take the form of ratios or 
proportions. A test for the independence of size (a volume measure- 
ment) and ratios of proportions has been derived. The test uses the 
multiple correlation coefficient, and tables of both conditional and 
tmconditional power of the multiple correlation coefficient have been 
prepared. The results are in preliminary manuscript form (Atta and 
Mosimann) . 

In connection with statistical analysis of proportions and ratios, 
conputer programs giving special covariance and correlation analysis 
and line plots were prepared. Some cluster methods were provisionally 
studied in this connection. 

79 



A new direction v/as undertaken with the stoody of M>d.8ls for population 
and individual growth. Since biomedical measuraiients o£ rates require 
observations at two discrete points o£ ti:i^, such Keasurements may or 
may not be directly comparable with the * 'instantaneous" rates used to 
determine ccmmon growth curves . Discrete time rates and Fodels ;, 
analogs to the conroon continuous models like the logistic and Gaispertz. 
were presented and studied. ITie discrete tirae models ai'e shcnvn to 
lend themselves well to computer siiKilation, 

In addition to the above study j irodels for growth of two populations, 
where one population consumes or destroys the other, were developed. 
Ihe models apply to the consunption (destruction) of a renewable 
resource by an expanding consumer population. Discrete and continuous 
time models were conpared in depth. Discrete time nodels lend them- 
selves readily to ccsiiputer simulation. Continuation time models gave 
somewhat neater mathematical results tlian their discrete counterparts. 
An invited talk on these topics was presented to the Computer Science 
Department of the University of Montreal. In connection with her 
Project Stride Training, Mrs. Clark prepared programs for this project 
for independent stxidy credits at American University. 

A time-sharing program for seaTxhing cataloged bacterial and viral 
mutant strains was prepared. Using this program, an investigator 
can iimiediately determine all cataloged strains satisfying or almost 
satisfying conditions he has prescribed. Personnel of the Laboratory 
of General and Conparative Biochemistry, NINfi, are now using the 
program as a research aid. 

The chemical reaction mechanisms study reported last year has been 
revised and submitted for publication. 

Proposed Course: 

Continued development of dispersion analyses for non-normal multi- 
variate distributions in biomedical data. 

Continued work on computer siimilation of population growth connected 
with bacterial growth and also immunological response reactions, 

A continuing DCRT project (principal investigator, J. M. S. Prev;itt, 
1MB) is that of the conputer screening of cytological and histological 
pictures for evidence suggesting papillary bladder tumors. For this 
project, various graph- theoretic techniques suitable for picture 
processing are being investigated. 

Publications: 

1 , Hutchinson , G. : The representation of lattices by modules . 
Bull. Amer. Math. Soc . , Vol. 79: 172-176, 1973. 



80 



2. Hutchinson, G.: Recxirsively unsolvable word problems of modular 
lattices and diagram- chasing. J. of Algebra , (in press) 

3. Hutchinson, G.: On classes of lattices representable by modules. 
Proceedings of the Houston Conference on Lattice Theory , (in press) 



81 



Serial No. DCRT 3.2 

1. Laboratory of Applied Studies 

2. Office of the Chief 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Evaluation of Con^juter-assisted Systems for Patient Care 

Previous Serial Number: Same ■ 

Principal Investigators: J. J. Bailey, M. Horton 

Assisted by: M. Douglas, M. Gelfand 

Co- Investigators: S. Epstein, S. Itscoitz, L. Grauer, J. Hirshfeld, CB, NHLI 

G. Johnson, E. Jones, M. Green, F. Allen, NM, CC 

H. Ostrow, CSL, DCRT 

J. Prevd.tt, EMB, DCRT 

Man Years: Professional - 3.5 
Other -0.3 

Project Descriptions: 

Two sub -projects are currently being conducted in the general area of the 
evaluation of health care systems, particularly those concerned with 
cardiovascular problems. These projects are: a) evaluation of conputer 
diagnosis of electrocardiograms (ECG) ; b) conputer-assisted studies in 
nuclear medicine. These projects were introduced in last year's report. 

A. Automated ECG Analysis 

Objectives: 

To assess on a sound statistical and clinical basis the usefulness of 
conputer programs for the diagnosis of electrocardiograms. ECG's from 
a range of cardiovascular disease, as well as presumed normal, are 
selected in cooperation with the Cardiology Branch, NHLI. 

Progress during FY 73: 

A clinical evaluation of three coii5)Uter programs was achieved in 1150 
ECG tracings and a preliminary result reported (1) . A full report has 
been con^jleted (2) . 

A refined technique for testing reproducibility of ECG programs was 
devised and preliminary results reported on the same three programs (3) 
These programs are: PHS version D, Mayo Clinic (1968), and IBVl 
(e3q)erimental 1971) . 



82 



In F\'' 72 only the Mayo Clinic program was in^lemented at NIH. 
Additional programs in^jlemented in FY 73 are: PHS version D, VA 
(Pipberger 1971) , and lEM (1972) . The JEM (1972) program has been 
selected as appropriate for routine use at NIH. Ilie output has 
already been reformatted so that it fits into patient charts. 

Proposed Course: 

Additional programs now being implemented are: PHS version E and 
VA (Pipberger 1972) . Continued evaluation and publication of the 
results is an important service to the consuming public perfoimed 
by NIH. In collaboration witJi the Cardiology Branch, NHLI clinical 
documentation of certain inportant diagnoses, namely ventricular 
hypertrophy and myocardial infarct, will provide a population for 
rigorous testing of the programs. 

A worthwhile by-product of this work will be the development of a 
technique to store ECG tracing (digital data), vector loop, and 
conq^uter output on microfiche. This is anticipated to be a cheap 
and very efficient solution to the storage problem now existing in 
the NIH ECG laboratory. 

Reports and Publications: 

1. Bailey, J. J., Hirshfeld, J. W. , Jr., Grauer, L. E., Karsh, R. B., 
and Itscoitz, S. B.: Unreliability of couputer programs in 
interpreting electrocardiograms. Circulation , Vol. 46: II-8, 
Oct., 1972. 

2. Bailey, J. J., Itscoitz, S. B., Hirshfeld, J. W. , Jr., 

Grauer, L. E., Horton, M. R. , and Karsh, R. B.: A comparison of 
three computer programs for electrocardiographic interpretation. 
Submitted to Circulation , Apr., 1973. 

3. Bailey, J. J., Horton, M. R. , and Pottala, E. W. : Sensitivity 
of ECG conputer diagnostic programs to digitizing techniques. 
Proceedings of the 2nd National Conference of the Society for 
Conputer Medicine , Nov., 1972. 

B. Computer Assisted Studies in Nuclear Medicine 

Ob j ectives : 

To provide computer-based mathematical analysis in support of 
diagnostic and research activities of the Nuclear Medicine Depart- 
ment, C. C. 



83 



Progress during FY 73: 

In FY 72 a joint LAS-CSL team wrote specifications, evaluated 
competitive proposals, and selected a mini-conputer system to be 
mated with two gamma scintillation cameras in the Nuclear Medicine 
Department, C. C. In FY 73 there was extensive checking of the hard- 
ware and the vendor- supplied software using data from known radio 
nuclide sources arranged in some fixed image (called a '"phantom") . 
Also in FY 73 a disk was added to che system and the necessary soft- 
ware changes are nearly co^leted. 

Extensive utility programs have been inplemented on the mini -computer 
system including manipulation of time functions, curve fitting, image 
enhancement, and functional mapping for dynamic studies (references 
1-3) . Programs for dynamic studies have found jjimediate use in 
clinical applications (references 4 , 5) . 

The main advantage of the mini-coiT5)Uter is its ability to handle the 
high rate of scintigraphic data and to capture it in computer proces- 
sable form. Its disadvantage is that sophisticated image processing 
programs are too large to be inplemented on it. Therefore an 
in5)ortant development was a method making scintigraphic data captured 
by the mini-conputer con5)atible with the large scale computers at 
DCRT. This has allowed images to be framed and displayed on the 
PDP-10 in 128 x 128 arrays; v/hereas the mini -computer can only handle 
64 X 64. 

Programs to do orthogonal transforms (Fourier, Hadamard, Slant) of 
images have been developed for the PDP-10. A collection of special 
phantoms has been captured on the mini -computer for the purpose of 
testing the use of orthogonal transforms in image restoration by 
deconvolution and non-restorative image enhancement. A collection 
of sophisticated image processing programs called PBEP is also 
available on the PDP-10 for scintigraphic applications. 

A hardware electrocardiographic trigger based on the R wave was 
designed by CSL and interfaced with the mini-conputer system. This 
will allow scintigraphic images of the heart at any desired time 
during the cardiac cycle which is important to the non- invasive 
estimation of cardiac performance. 

Proposed Course: 

Software and hardware development will continue along the indicated 
lines, with enphasis on clinical applications. An additional scope 
is being purchased so that the clinicians in the Nuclear Medicine 
Department will have quicker access during the daily routine readings 
to the results of conputer- aided studies. 



84 



Reports and Publications: 

1. Larson, S, M. , Bailey, J. J., Schall, G. L. , Griffith, J. M. , 
and Johnston, G. S.: Radioisotope technique for measuring 
regional organ blood flow. International Journal of Applied 
Radiation and Isotopes , Vol. 23: 390-392, August, 1972. 

2. Larson, S. M. , Millar, R. C. , Chalmers, T. C., Kramer, R. T., 
Bailey, J. J., and Johnston, G. S. : Quantitation of hepatic 
blood flow by xenon clearance: a primate model. In The Liver , 
Proceedings of the First GSTAAD, Sept., 1972, G. Paumgartner 
and R. Preisig, editors. S. Karger Publishing Co., Basle, 
Switzerland. 

3. Gelfand, M. J., Green, M. V., Beiiley, J. J., and Johnston, G. S.: 
A utility program package for manipulating conputer- acquired 
static and dynamic radionuclide image data. 20th Annual Meeting 
of the Society of Nuclear Medicine, June, 1973. 

4. Gelfand, M. J., Allen, F. H. , Johnston, G. S., and Bailey, J. J.: 
Functional mapping of the renal intake and excretion of Iodine 
(131) - lodohippuran, 20th Annual >feeting of the Society of 
Nuclear Medicine, June, 1973. 

5. Kramer, R. J., Roberts, A. J., Gelfand, M. J., Milder, M. S., 
Hall, C. A. , and Mcintosh, C. L. : Ventilation Perfusion Scinti- 
photography in evaluating the pre-operative cardiac patient. 
Submitted to Surgical Forum . 



85 



Serial No. DCRT 3.3 

1. Laborator}' o£ Applied Studies 

2. Office of the Chief 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Computer Methods in Cardiologic Research 

Previous Serial Number: Same 

Principal Investigators: J. J. Bailey, M. Stroot, M. Gelfand 

Assisted by: M. Douglas, W. Scott 

Co -Investigators: S. Itscoitz, CB, NHLI 
J. Mason, NASA 

Man Years: Professional - .75 
Other - .5 

Project Description: 

Ob j ectives : 

To test and compare the reliabilities of proposed indices of 
myocardial contractility. Such indices include isometric 
contraction time, systolic ejection time fraction, mean systolic 
pressure, maximum change in pressure (max dp/dt) during isometric 
contraction, maximum dp/dt divided by the pressure (p) at which 
it occurs and maximal velocity of the contractile element (Vmax) , 
(e.g., the Sonnenblick or Mason indices). From such parameters 
may be computed indices related to ventricular power or the 
efficiency of power conversion. 

Progress during FY 73: 

Seven indices of left ventricular performance were assessed in 
46 patients. Five had normal ventricles and the rest had disease 
of coronary vessels and/ or cardiac valves . (See report) . 

The measurements were done as a part of the routine cardiac 
catheterization procedure and required only the insertion of 
the catheter-tip manometer into the left ventricle. The data 
was recorded on FM analog tape and processed by the CDC 3100 
hybrid computer. 



Proposed Course: 

The methods will be used in the right ventricle o£ patients 
with mitral valvular disease to evaluate their pre-operative 
status and estinate their post -operative prognosis. 



Reports : 



Gelfand, M. J., Clancy, D. L.: Comparison o£ indices of 
left ventricular (LV) performance in man. Circulation 
46(4): 11-55, Oct. 1972. 



Serial No. DCRT 3.4 

1. Laboratory o£ Applied Studies 

2. Office of the Chief 

3. Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Statistical Research in Clinical Pathology 

Previous Serial Number: Same 

Principal Investigators: Eugene K. Harris, David DeMets 

Assisted by: Willie Mae Scott 

Co- Investigators: D. Young, Clinical Chemistry, CC 
F. C. Bartter and staff, EB, NHLI 
J. D. Decker, A ^ RB, NIAMD 

Man Years: Professional - 1.5 1 

Other - 1.0 

Project Description: 

Background: 

The studies of variation in normal blood chemistries which form the 
backgroimd of this project have been anply discussed in earlier 
annual reports. 

Progress during FY 73: 

Based on published work described in the previous report, methods 
have been developed, programmed, tested and published for the trans- 
formation of frequency distributions of clinical chemistry measure- 
ments to the symmetrical Gaiossian form. This allows simple but 
precise estimation of ranges and cumulative proportions, useful in 
routine diagnosis. A study of co-variation among connnon blood 
constituents (e.g., calcium-protein, albumin-globulin) has led to a 
definition of the statistical problems and assunqjtions underlying 
correlations typically calculated from collections of single-saiiple 
(one per person) data. 

Itore recently, attention has been focused on possible applications 
of time series analysis to repeated measurements of blood chemistries 
during intensive, evaluations of drug therapy in patients. This 
theory, based on the correlations among observations over time, 
should lead to new design protocols and much more powerful analytic 
techniques in such inpatient studies. Applications to the study of 
continuous or very frequent non- invasive measurements in normal 
volunteers are also expected during the coming year. 



Publications: 



1. Harris, E. K. and DeMets, D. L.: Estimation of normal ranges 
and cumulative proportions by transforming observed distri- 
butions to Gaussian form. Clinical Chemistry . 18: 605, 1972. 

2. Harris, E, K.: Estimating correlations from single-sample 
distributions of measurements. Clinical Chemistry. 19: 81, 
1973. 



89 



Serial No. DCRT 3.5 

1. Laboratory o£ Applied Studies 

2. Applied Mathematics 

3. Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Applied Mathematics Section 

Previous Serial Number: Same ■ 

Principal Investigators: J. Fletcher, J. Ashbrook, E. Hill, J. Rinzel 

Co- Investigators: A. Spector, Univ. of Iowa (formerly with Lab. of Metabolism) , 
NHLI; B. Hubbard, Univ. of Maryland; R. Shrager, PSL, DCRT; 
C. Lenfant, NHLI; R. Sinpson, Lab. of Biophysical Chem. , 
NIAMDD; H. Hoffman, Biometry Branch, NICHD. 

Man Years: Professional - 3.08 

Project Descriptions: 

Background and Objectives: 

The primary responsibility of the Applied IVfethematics Section is to 
provide DCRT and the NIH with high level mathematical competence for 
biomathematical modelling and data analysis. This competence includes 
both theoretical and applied techniques, as well as numerical computa- 
tion methods. Each indiA/ldual in the section has a primary specialty 
in con5)uter science or mathematics, and in addition, each is a capable 
programmer . 

Progress during FY 73: 

1) Modeling of Biological Diffusion Processes: 

The conceptual models developed during FY 71 and 72 have now been 
generalized to include nonlinear effects. The discovery of the exact 
form for the inverse of the Adair formulation of the oxygen-hemoglobin 
dissociation function has enabled a consideration of the conplete 
tissue-oxygen kinetics in the full time dependent model. New approxi- 
mations have widened the applicability to second order kinetics in 
metabolism and a consideration of time dependent flows. Simulations 
of many in^jortant cases have been carried out, and the results are 
being published. 



90 



2). Macroniolecule-Ligand Binding Madels: 

Techniques for the analysis o£ experimental raacroniDlecule -ligand binding 
data have been analyzed, both theoretically and experimentally. New 
procedures for determining the nature and extent of binding are being 
standardized and conqjuter programs to implement these techniques are under 
development. The resiiLts have been described both in conferences and in 
publications , 

A study of the binding of medium-chain fatty acid to human plasma albumin 
has been coiipleted and published. The partition of long-chain fatty acids 
between n-heptane and an aqueous buffer is currently being investigated. 

3) . General Support Activities : 

a. M3DELAIDE: 

Because of changes in System 370 software support, extensive revisions 
to MDDEIAIDE are required. A cooperative project with CCB and PSL is 
underway to modify >foDELMDE so that the system may be invoked from a 
graphics terminal under the new Time Sharing Option (TSO) . Revisions 
are expected to require at least one year or more. 

b. Integrated Plotting Package: 

A groi^) of gr^hics routines, obtained from the University of Chicago 
and the National Bureau of Standards, have been iiqplemented on the IWl 
System 360/370. The Integrated Plotting Package (IPP) may be en^jloyed 
to generate inteimediate (or neutral) text which will be processed, 
possibly after storage on an external file, to some plotting device. 
As in^lemented at NIH, the IPP routines may be used to generate plots 
on (1) a CalComp plotter, (2) an SD4460 microfilm/microfiche unit, 
and (3) a line printer. Documentation for the package is being 
prepared. 

c. Low Weight for Age Study: 

The Integrated Plotting Package is being enqjloyed to generate contour 
plots for a Low Weight for Age study. The primary investigators are 
NICHD. 

d. Neurological Modeling: 

A Simulation Model of Nerve-Cell Functioning, based on the model 
constructed by Donald H. Perkel has been modified to use random 
access storage and is running on the PDP-10. An algoritlim designed 
to analyze Neurological -Spike data is now coii5)lete and is being 
evaluated for publication. 



91 



4J . ftothematical Curve Fitting ^fethods : 

Alternatives to least squares methods are being considered for data 
analysis. The Chebyshen Minimax method of fitting data is currently under 
study for purposes of in5)roving the exchange algorithms. 



5) . Publications : 

A 

1. Spector, A, A., Ashbrook, J. D, , Santos, E. C, and Fletcher, J. E.: 
Quantitative analysis of free fatty acid uptake by mammalian cells: 
Laurie acid and human erythrocytes. J. of Lipid Research , Vol. 13, 
No. 4, July 1972. 

2. Ashbrook, J. D. , Spector, A. A., and Fletcher, J. E.: Medium chain 
fatty acid binding to human plasma albumin. Journal of Biological 
Chemistry , Vol. 247, No. 21, Nov. 10, 1972. 

3. Fletcher, J. E.: On the effects of time -dependent flow in capillaries 
si^plying substrates to living tissues . SLAM Review (Chronicle) , 
Vol. 15, No. 1, January 1973. 

4. Fletcher, J. E., Ashbrook, J. D. , and Spector, A. A.: Computer 
analysis of drug-protein binding data. (Invited paper, to appear in 
the proceedings, 1973). Annals of the New York Academv of Sciences, 
1973. 

5. Spector, A. A., Santos, E. C, Ashbrook, J. D., and Fletcher, J. E.: 
Influence of free fatty acid concentration on drug binding to plasma 
albumin. (Invited paper, to appear in the proceedings, 1973). Annals 
of the New York Academy of Sciences , 1973. 

6. Fletcher, J. E.: A mathematical model of the unsteady transport of 
oxygen to tissues in the microcirculation. Advances in Medicine and 
Biology . Proceedings of the Third International Synposium on Oxygen 
Transport to Tissue. Plenum Press, (to appear in Summer, 1973). 



A 

Listed as "in press" in- FY 72 report. 



92 



Serial No. DCRT 3.6 

1. Laboratory of Applied Studies 

2. Office of the Chief 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Siinulation and Modeling in Neurophysiological Research 
Previous Serial Number: Same 

Principal Investigators: James A. Mortimer (Project A) 

Erik Pottala (Projects B and C) 

Co-Investigators: E. B. Evarts, LNP, NIMH (Project A) 

T. R. Colbum, IR, NIHM (Projects B and C) 

R. G. Justiss, CSL, DCRT (Project C) 

D. Mirvis, LPP, NINDS (Project C) 

D. Wasserman, NIOSH, Cincinnatti, Ohio (Project C) 

J. Borer, CB, NHLI (Project C) 

B. Sharma, University of Nkryland (Project B) 

D. Humphrey, Emory University, Atlanta, Ga. (Project B) 

Man Years: Professional - 1.5 
Other - 1.0 

Project Description: 

A- ^o^^^^^°"^-^ ^°^^^ ^"^ E xperimental Study of Mammalian Cerebellar 
Objectives: 

1) To evaluate alternative hypotheses of cerebellar function 
through study of the electrophysiological activity of the 
monJcey cerebellum in relation to motor activity. 

2) To relate the physiological findings of this study to 
results obtained by computer simulation of a cerebellar 
network model. 

3) To develop single cell and network models for the mammalian 
cerebellum, which will permit direct comparisons to be made 
between simulation output and physiological data 



93 



Progress during FY 73: 

During this year an additional experiment was carried out 
to clarify the relationship between changes in cerebellar 
activity and motor reactions to \^h.ich they were related. 
The results o£ this experiment as well as previously- obtained 
findings have been reported (See Publications) . 

In the course of the physiological experiments described in 
this and the previous annual report, several new findings 
related to the mode of operation of the cerebellum in normal 
primates have been obtained. In particular, it has been 
shown that, relative to a specific motor reaction, the two 
input pathways to the cerebellum provide fundamentally 
different types of information. The mossy fiber input seems 
to be related to the patterning of muscular activity related 
to the motor task, whereas the climbing fiber input is better 
related to the nature of the stimulus. Furthermore, these 
experiments provide evidence that disinhibition of cerebellar 
nuclear cells as a mechanism of cerebellar control of movement 
is much less important than had been previously assumed. The 
composition of cerebellar output is shown to depend substan- 
tially upon the inhibitory input from Purkinje cells as well 
as the excitatory input to cerebellar nuclear cells from the 
mossy fibers. 

In addition to the physiological experiments, two theoretical 
papers have been written dealing with cerebellar operation 
upon the analysis of computer simulation output from a cere- 
bellar network model. 

A third project has been initiated in collaboration with 
Erik Pottala (LAS/DCRT) , in which the influence of dendritic 
spikes on the input-output relations of single Purkinje cells 
is studied by means of a hardware neuron model. 

Proposed Course: 

During the next year, the single cell simulation project 
will be continued. It is expected that this simiolation will 
yield valuable information regarding the importance of active 
membrane properties of dendrites in the integration of 
realistic spatio-temporal patterns of synaptic input. In 
addition to the single cell model, an expansion of a larger 
network model is anticipated, which will incorporate the 
results of the single cell simulation. 



94 



Publications: 

1. Mortimer, J. A.: A computer model of mammalian cerebellar 
cortex. Computers in Biology and Medicine (in press) . 

2. Mortimer, J. A.: Temporal sequence of cerebellar Purkinje 
and nuclear activity in relation to the acoustic startle 
response. Brain Research 50: 457-462, 1973. 

3. Mortimer, J. A.: Influence of barbiturate anesthesia on 
cerebellar operation: A computer simulation study 
(submitted to Kybernetik) . 

4. Mortimer, J. A.: Response of cerebellar neurons in monkeys 
to startle-producing stimuli (in preparation) . 

Hardware Neural Mjdeling 

Objectives: 

The objective of the hardware neural modeling study is to 
investigate small neural nets through the use of physiologically 
realistic hardware neural models which incorporate a distri- 
buted input system (analogous to a dendritic net) and are 
able to simulate action potentials. 

Progress during FY 73: 

n-ie mo&el has given insight into the manner in which an action 
potential can modify the shape and duration of post-synaptic 
potentials and the spatial -temporal interactions amoi.g such 
potentials. A preliminary study of the steady-state input- 
output firing characteristics of the model were presented in 
part at the 24th Annual Conference on Engineering in Medicine 
and Biology. The full study has been published (reference 1). 
The hardware implementation of the electronic compartmental 
neuron was totally redesigned in FY 72 and is now interfaced 
to the MAC-16 computer, which allows an investigator to liave 
interactive computer control and facilitates rapid simulation 
of neural nets with varying branches, conduction times, and 
excitatory or inhibitory synaptic connections (references 1-3). 
This neural modeling system has been used to simulate the 
walking reflex. (Study formed a basis for a Ph.D. thesis at 
University of Maryland Department of Electrical Engineering) 
(reference 4) . 



95 



Proposed Course: 

1. Purkinje Cell Simulation 

Co -Investigator: J. Mortimer, LAS/DCRT 

The neuron models will be interconnected to investigate 
dendritic spike activity and other characteristics 
attributed to Purkinje cells, (cerebellum) 

2. Reciprocal Inhibition-Excitation Behavior 

Co- Investigator: D. Humphrey, Emory Univ., Atlanta, Ga. 

Small net simulations with various internal inhibition 
and excitation properties. 

Reports and Publications: 

1. Pottala, E. W., Colbum, T. R., and Humphrey, D. R. : 

A dendritic compartment model neuron. IEEE Transactions 
Biomedical Engineering , Vol. ME-20, Num. 2, p. 132, 
March, 1973. 

2. Pottala, E. W. : Computer control of a compartmental 
model neuron. Medical Instrumentation , Vol. 7, No. 1, 
p. 74, January, 1973. 

3. Pottala, E. W.: A hybrid neural simulation system. 
Proceed. 1973 Summer Computer Simulation Conference , 
July, 1973. 

4. Sharma, B. , Pottala, E. W., Zajac, F. E. III.: Analysis 
of the walking reflex via a compartmental model neuron. 
Proceedings of the 25th American Conference for Engineering 
in Medicine and Biology , p. 67, 1972. 



96 



C. MAC- 16 Research System Development (the LAS Computer) 

Backgroimd: 

A mini-conputer has a place as a research tool i£ it has a variety 
o£ interfaces to control different types of peripheral devices and 
if it has an easy-to-use research programming system x^/hich elimi- 
nates the need of the researcher to know the fine points of these 
various peripherals. To obtain these desired hardware and soft- 
ware systems is the objective of this project. 

Progress during FY 73: 

In FY 72, the MAC-16 computer was interfaced to the Marquette 
tape drive (for analog ECG's), the Honeywell 7600 Analog Tape 
Transport, the neural control panel, and the general purpose 
switch-filter network. In FY 73 additional hardware was added 
to the system including a real time spectrum analyzer and 
ensemble averager. 

Software development in FY 73 has enabled the MAC- 16 system to 
become a versatile tool for handling physiological data 
(references 1 and 2) . With this system the investigator can 
easily edit huge amounts of continuously recorded data; he can 
easily find the frequency content of the significant data as 
well as that of any interfering noise; he can select the optimal 
analog filter and digitization rate; and he can arrange for 
automatic pre-processing - editing, filtering, and digitizing -- 
of his data so that he can get the most economic use of a large 
scale digital conputer. 

In FY 73 these advantages were demonstrated with cardiovascular 
data recorded on live monkeys which enable the investigators to 
study myocardial contractility (reference 3) , (Also demonstrated 
in ECG reproducibility study, see project 3.2). 

Proposed Course: 

Continued applications will be explored: 

1. Real time spectral analysis of accelerometer and electro - 
myogram (EMG) data to assist in evaluation of Parkinsonian 
patients and their response to drugs. 

2. Spectral analysis of vibrating equipment and its effect on 
human EMG conducted in collaboration with the National 
Institute of Occupational Safety and Health (NIOSH) . 

3. Analysis of data from real and simulated cerebellar Purkinje 
cells . 



97 



Publications : 



1. Pottala, E. W. , Justiss, R. G., Horton, M. R. , and Bruce, M. C: 
Multipurpose computer system for biological research. Proceedings 
of the 25th American Conference for Engineering in Medicine and 
Biology , p. 105, 1972. 

2. Pottala, E. W. : Con^uter controlled real time spectral analysis 
as a tool for processing physiological data. Medical Instrumenta- 
tion , Vol. 7(1): 74, Jan., 1973. 

3. Mirvis, D. M. , Koph, G. S., and Pottala, E. W. : The effect of 
low frequency response catheter systems on the calculation of 
force velocity parameters in the intact heart. Proceedings of 
the 10th Annual Rocky Mountain Bioengineering Symposium , May, 
1973^ 



98 



July 1, 1972 through June 30, 1973 

PUBLIC HEALTH SERVICt - NATIONAL INSTITUTES OF HEALTH 
DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 

Summary of Branch Activities 1. DCRT - 4 

Serial Number 

2. COMPUTER SYSTEMS LABORATORY 3. Alan Demmerle 

Chief 

This marks the first year in which the Laboratory's programs are not reported 
on individually^ but instead are abstracted in this Laboratory Summary. This 
step has been taken in order to emphasize the complex, team-related aspects 
(as compared with individual research efforts) of our projects; however, the 
name of a cognizant project leader, who may be contacted for further details, 
is listed in parenthesis with each project abstract. 

The mission of this Laboratory is to identify and seek solutions for problem 
areas in biomedical research and clinical care in which the computer promises 
improved research productivity, or improved care. We limit ourselves to areas 
unsuited to solution on the DCRT central computer facility. The central com- 
puter facility is unsuitable for applications requiring real-time data collec- 
tion, analysis, display, and experimental control, where economic consider- 
ations favor a small computer, or where equipment proximity is important. 

Our general mode of operation is to analyze requirements, specify and design a 
complete system including both equipment and software components, to satisfy 
these requirements, purchase those parts of these systems which are commercially 
available and design and build, in-house, those parts of the system which are 
not conmercially available. This activity requires concentrated team effort 
on the part of engineers, mathematicians, programmers and systems analysts. 
Such projects normally require several years from inception to completion. 

Laboratory Systems 

Among the large systems which we have been working on for the past several years 
are one in NIDR (Syed), one in NIMH (Syed), and one in NIAMDD (Shapiro). During 
the past year we have completely withdrawn from work on the NIDR system. During 
this fiscal year we assisted NIDR with electrically interconnecting this system 
over telephone lines (which involves program development as well as hardware 
interconnection) to the DCRT central facility and since the system is now oper- 
ational in accord with their needs, there is no need for further development 
on our part. During the past year we have been working on the NIMH system to 
develop an interactive family-interaction experiment with an NIMH scientist 
and ve have begun to interconnect this system with the DCRT central facility 
to provide for more complex analysis of some of the data which NIMH scientists 
collect on the system. We expect to complete this project in the next fiscal 
year. Both the NIDR and the NIMH have a small staff to man their computer sys- 
tems for minor development efforts and software maintenance. We think of the 
NIAMDD system (Building 2) as a more long term commitment than the other two 
systems. This systern now has seven instruments interfaced to it, CARYS 118, 

99 



60, 14, and Beckman IR7 spectrophotometers, a computer of average trans i e'^ts , 
PDP-8 and a Raytheon 704 of which one spectrophotometer, the GARY 118, was 
interconnected during FY73. The work on the GARY 118 represented a significant 
research accomplishment for our laboratory. A number of modes of operation 
were designed into the interface of the instrument and the repetitive scan 
feature which we designed allows the users (Drs. Miles and Howard) to run many 
runs at different temperatures without any intervention on their part. This is 
a feature which many users of the GARY 118 would like to have but which is not 
available from the manufacturer. We have also been working toward interconnec- 
ting this system via telephone lines to the DGRT central facility, but have not 
completed this. The purpose of such interconnection is to support large statis- 
tical analyses which are better handled on the central facility due to the 
magnitude of the data processing and the availability of standard statistical 
routines. Our level of effort on this project has been cut back significantly 
over the past year but we expect over the next year to devote about two man- 
years to the NIAMDD system. 

During the past year we have expended a significant effort on three smaller 
laboratory systems and worked less intensely on the installation and hardware 
interfacing efforts on two others, those belonging to Drs. Felsenfeld and 
Hagins which are now almost complete. The small system supporting Dr. Hagins 
research into the structure and function of the retina, has worked out partic- 
ularly well. Because of the special features built by CSL to allow for control 
of the experiment and the interconnection of the small computer with the larger 
Honeywell computer in the building, the support given by computers to this 
research has been noteworthy. The time to complete a set of experiments has 
been reduced from weeks to a few hours, but most importantly, Dr. Hagins is 
now planning and performing experiments which, without this successful com- 
puter support, would have been impossible. The three projects requiring a 
significant effort are for Dr. Yoshikami, NEI (Schultz), Dr. Waxdal (Shapiro), 
and for the Pulsed NMR Spectrometer NIAMDD, Dr. Becker (Colburn). The NEI 
system is intended to collect data from experiments on animal retina and 
through averaging and correlation techniques, analyze and display tliat data. 
The computer equipment to do this has now been specified and a RFP was issued 
in April 1973 to purchase the bulk of the system. Interfacing this system 
to the laboratory equipment and developing the software for it will be done 
during the coming one and one-half fiscal years. Dr. Waxdal is concerned with 
biochemical research using two Beckman amino acid analyzers and a Gilford U,V. 
spectrophotometer. We specified and procured the elements of a system to 
accommodate his needs. The first purchased computer equipment was deficient 
and so the contract was terminated and the equipment ordered from the second 
choice among the original bidders. This equipment is to arrive in early FY74. 
We expect to expend about. another two man-years developing this system using 
the efforts of three people. The pulse NMR Spectrometer system utilizes a 
Raytheon 704 computer with a BEIB built RF pulse system. The computer collects, 
averages and transforms data from the time to the frequency domain. The sys- 
tem has become operational at expected performance level during FY73. Another 
two man-years will be expended during FY74 to extend and enhance the system 
for larger data tables, more finished spectrum outputs and more efficient 
operation. 



100 



CSL has worked on five systems in support of clinical care: 

1. The Intensive Care Unit Computer System (Syed) has been developed for 
the NHLI. During most of the year six people have been working full 
time on this system and have now fully implemented the system for one 
patient in the Heart surgery recovery area. The system now collects, 
analyzes and displays data from EKG, temperatures, fluid loss (urine 
and blood) and arterial and venus pressure. It also maintains and dis- 
plays admission data on that patient. Displays of this analyzed data 
are available at the bedside. During the coming fiscal year we will be 
adding inputs from other beds to the system. We expect to sustain this 
project with the efforts of four people working full time for the coming 
fiscal year. 

2. The existing Clinical Pathology Computer System has been a source of 
long standing dissatisfaction due to an array of difficulties arising 
from the lack of responsiveness, adaptability, maintainability and 
reliability of the system as a whole. We were requested (Plexico) 

early in this fiscal year to provide assistance to the Clinical Pathology 
Department in two areas, the first being to overcome some of the most 
pressing problems in order that the system could continue to function 
in a reasonable manner, at least for the immediate future. The second 
area was to identify a long term approach by which the Clinical Path- 
ology Department will be able to resolve the aforementioned problems 
from a total system point of view. 

The most serious problem areas, and those identified for a short range 
solution, were due to gross unreliability and lack of reasonable main- 
tenance service for equipment in the Chemistry and Hematology labora- 
tories which serves to enter laboratory results into the computer sys- 
tem. To solve this problem we have procured two DEC PDP-11 computers 
and a number of Hazeltine cathode ray terminals. This equipment, 
together with special terminals we have developed for the interfacing 
of two automatic instruments will serve, in conjunction with extensive 
software written by CSL, as the new laboratory data entry system. These 
new systems will be functional by the summer of 1973 and will enhance 
as well as replace the functions of existing equipment. We expect, how- 
ever, to continue our programming efforts for another several months in 
order to further enhance and increase the versatility of these systems. 
This effort has, to date, required about one and one-half man-years of 
effort. 

During the latter quarter of this fiscal year, CSL, along with members of 
the Clinical Pathology Department, and representatives of the Office of 
Clinical and Management Systems of the Clinical Center have been examining 
the alternatives in regard to an entire overhaul of the Clinical Pathology 
Computer System. The most likely alternative involves the purchase of a 
system which is already developed and in use in other hospitals but would 
nevertheless require some modification to maximize its performance in the 
NIH environment. We expect this to be one of our major projects in the 
coming fiscal year. 

3. The Pulmonary Physiology Laboratory and CSL (Plexico) have been working 
toward the automation of many of their clinical and research protocols. 
This automation is expected to relieve some of the medical technician 

101 



workload and facilitate more rapid data analysis and thus allow for 
modification of clinical and research protocols while the patient is 
still there at the site of the testing equipment. The computer, Data 
General Nova 1200, originally purchased by NHLI as an integral part of 
one of the laboratory instruments, has been expanded with more core and 
a disk and it is now being interfaced to three more instruments (medical 
mass spectrometer, wedge spirometer and oscillating resistance unit). 
Work will continue during the coming fiscal year to interconnect more 
pulmonary testing instruments. 

4. CSL has worked jointly with LAS, DCRT and the Nuclear Medicine Department 
of the Clinical Center in the development of a Pho Gamma Camera computer 
system (Schultz). CSL has contributed the engineering expertise required 
for the acquisition and development of this system and is now considering 
additional system requirements and the interconnection of a new imaging 
device. 

5. A number of small hardware devices to augment computer analysis of bio- 
medical signals have been developed by CSL over the past several years. 
This type of work continues. We are currently modifying earlier models 

of two devices, one which analyzes EC6 signals and one for aortic pressure 
signals external to the computer, thereby saving computer capacity for 
other things. When completed these devices will become part of the ICU 
System. We have developed a device which allows for the easy collection 
of ECG data over a telephone. It is necessary only to place two fingers 
on two CSL designed concentric electrodes and using a telephone, transmit 
an ECG of sufficient quality to assess heart rate. Forty copies of this 
device were made by George Washington University Hospital and are used 
to check the pacemaker rate from patients while in their homes, thus 
saving them trips to the G.W. Clinic. 

We are also working on a processor of phonocardiograms for the eventual 
computation of pre-ejection periods when used in conjunction with ECG 
and pressure derived data. 

6. For the past several years we have been developing what we have called a 
Medical Telecommunication System. This system was to demonstrate the 
feasibility of using a computer to aid the physician in the conduct of 
his practice with diagnosis, therapy planning and poison identification. 
The physician would use the touchtone telephone as input to the computer 
and the computer would respond in English over the telephone. 

The prototype has been completed and has demonstrated itself to be con- 
venient, easy to use, timely and economical. While we have achieved a 
substantial degree of success with the purely technical aspects of this 
project, progress toward establishing a meaningful field trial has been 
less notable. We have recently been contacted by the University of Colo- 
rado Medical Center concerning the possible use of the system in a poison 
control application, but it appears unlikely at this time that it will 
come to fruition due to their lack of funds. (They had hoped for RMP 
funds but RMP is now on its way out of business.) If no applications 
present themselves in the first few months of FY74, we will abondon 
the project. 



102 



other research activities of this Laboratory include the application of pattern 
recognition and cluster analysis techniques to drug data and other data from 
chemical compounds (Shapiro). Mass spectra data is being used as a basis for 
predicting the activity of drugs. This has been done in collaboration with 
the NHLI and HL, DCRT. We have also just begun studies of the metabolic path- 
ways of xenobiotics using a computer program designed to aid with the synthesis 
of organic molecules. This program is being modified to allow the pathways of 
metabolic reactions. The intent is to use the computer techniques in collabo- 
ration with researchers of NCI to get a clearer understanding of the metabolism 
of anti -cancer aaents. 



IO3 



July 1, 1972 through June 30, 1973 
PUBLIC HEALTH SERVICE - NATIONAL INSTITUTES OF HEALTH 
DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 

Summary of Branch Activities 1. DCRT - 5 

2. PHYSICAL SCIENCES LABORATORY 3. Dr. G. H. Weiss 

I . OBJECTIVES 

Tlie Physical Sciences Laboratory is devoted to the study of problems in 
physics and chemistry that relate to the biological sciences. Several 
disciplines are represented in the membership of the laboratory. These 
include applied mathematics, theoretical chemistry, and theoretical physics. 
Whenever possible the theoretical studies are performed in conjunction with 
experimental work, either in collaboration with workers in outside units, 
or by members of the Physical Sciences Laboratory working in other labora- 
tories at NIH. In addition to performing research of its own choosing, 
members of the Physical Sciences Laboratory provide consultation to other 
researchers at NIH on different topics in the disciplines represented in 
the Laboratory. These services are enumerated in the project reports. 

II. CURRENT LABORATORY PROGRAMS 

Summary 

1. We have begun to develop a theory of resolution in chromatography, 
particularly studying inhomogeneous systems for v^ich no theory has so far 
been available. Results obtained so far indicate that there is no particu- 
lar advantage to using a gradient in transient one dimensional chromatography. 
We are presently investigating two dimensional gel electrophoresis which is 

a practical system, and in v^ich there may be some advantage to having a 
gradient . 

2. The theory of van der Waals forces has been extended to include the 
effects of salt solutions. These turn out to have a large screening effect 
and change the forces between particles by a significant amount. Since 
most biological processes occur in the presence of salt solutions, the 
physical understanding of cellular interactions will be modified by these 
findings . 

3. Precise measurements of bacterial motility have been made by laser 
scattering. These measurements will be used to study chemotactic phenomena 
both in bacteria and in the study of white cells. 

4. The theory of clinical trials has been extended to include the effects 
of observational delay. It has been shown to be possible to bias the 
allocation of conpeting treatments so that fewer patients receive poorer 



105 



ones, while retaining the possibility o£ statistically discriminating 
bet\'/een the treatments. 

5. A 55 MHz pulse NMR spectrometer has been constructed and programs have 
been written and debugged to perform Fourier transform spectroscopy. 



106 



Serial No. 5.1 

1. Physical Sciences Laboratory/- 

2. Not Applicable 

3 . Bethesda 

PHS-NIH 
Individual Project Report 
July 1, 1972 through June 30, 1973 

Project Title: Theory o£ Biochemical Separation fechnique 

Previous Serial Number: 5.1 

Principal Investigator: George H. Weiss, Ph.D. 

Other Investigators: None 

Cooperating Units: David Yphantis, Ph.D., University of Connecticut, 
Gary Ackers, Ph.D., University of Virginia, and 
David Rodbard, M.D., Reproduction Research Branch, 
NICHD. 

Man Years 

Total : 0.3 

Professional : 0.3 

Other: 0.0 

Project Description: 

Objectives: 

To determine the physico-chemical effects influencing different 
biochemical separation systems such as ultracentrifugation, chromatography, 
and electrophoresis. To determine the quantitative significance of these 
effects. To devise numerical techniques for processing data from chemical 
separation procedures to determine properties such as molecular weight and 
diffusion coefficients. 

I^thods: 

Numerical and analytical solutions to linear and nonlinear 
partial differential equations. 

Major Findings: 

Approximate egressions have been developed for the noments of 
the concentration and elution profiles in inhcmogeneous chroimtography. 
These have been used to study properties of such systems related to resolv- 
ing power. Specific applications have been made to electrophoresis in 
polyacrylamide gels„ 



107 



We have devised a technique for using the time dependent information 
in isoelectric focussing to find moleaiLar para^neters . Heretofore iso- 
electric fociissing has been losed only as an equilibrium system, but the use 
of transient information gives one the capability of measuring diffusion 
and velocity parameters separately. 

An extensive numerical study of the effects of pressure on the 
association constant of polymerizing systems in the ultracentrifuge was 
concluded. Tne nwierical study suggested qiiite accurate approximations 
for taking into account pressure effects in the interpretation of such 
e-xperiments . 

A study was begun on optimizing gel parameters in two dimensional 
gel electrophoresis. 

Significance to Biomedical Research: 

A central problem in biochemistry is the separation of proteins. 
In the past chromatographic methods have been used for this purpose produc- 
ing crude and qualitative results. Presently electrophoresis and gel pore 
chronHtography are being developed as analytica], tools with some advantage 
over centrifugation. Analysis of these tools with a view towards the 
development of numerical methods for processing results of chemical 
separation experiments will lead to extremely accurate nethods for character- 
izing inportant protein properties. 

Honors and Awards: None 

Publications: 

Weiss, G. H. , Dishon, M. : On the solution of linear chemical 
separation equations with weak diffusion effects. Separation Science 7, 
585-599 (1972). 

Weiss, G. H. Ackers, G. K.: Magnitude of a finite equilibration 
effect in analytical gel chromatography. Biopolymers 11 , 2125-2130 (1972). 

Weiss, G. H., Dishon, M.: Approximate solutions of chemical 
separation equations with diffusion. Advances in Chemistry (to appear) . 

Weiss, G. H. , Dishon, M. : Resolution in non uniform chronHtography. 
Separation Science (to appear) . 

Rodbard, R. Chrambach, A,, Weiss, G. H.: Optimization of resolution 
in analytical and preparative polyacrylamide gel electrophoresis. Electro - 
phoresis and Isoelectric Fooissing in Polyacrylamide Gel Electrophoresis 
(to appear) . 



108 



Serial No, 5.2 



1 . Physical Sciences Laboratory 

2. Not Applicable 

3. Bethesda 

PHS-NIH 
IndiATidual Project Report 
July 1, 1972 through June 30, 1972 

Project Title: Theory o£ the Helix-Coil Transformation o£ Polypeptides 
in Solution 

Previous Serial Number: 5.2 

Principal Investigator: James A. Ferretti, Ph.D. 

Other Investigators: Robert Jemigan, Ph.D., George Weiss, Ph.D. and 
Julie Milste in, Ph.D. (NCI). 

Cooperating IMits: 

Man years: 

Total 0.6 

Profess ional : 0.6 

Other: 0.0 

Project Description: 

Ob j ectives : 

To elicit the molecular details of the kinetics of helix to 
random coil transitions in polypeptides. A further aim is the conprehension 
of the action of proteolytic enzymes of polypeptides. 

Methods : 

A better model to describe the conformational kinetics has been 
obtained. Only two molecular rates are required in this model. It permits 
a conplete realistic treatment of the process in terms of two states, helix 
and coil . Also we are treating the random coil configurations through the 
transition in more detail. In this case, transitions between additional 
regions of configurat ional space are permitted. 

Major Findings: 

Mean relaxation times have been calculated. Conparison of these 
values with experimental results permits a determination of the unit trans- 
ition time. Significantly this time is several orders of magnitude larger 
than had been anticipated by other with cruder models. Also, this time is 
aonsistent with our earlier interpretations of measurements. Our average 



109 



relaxation time is not specific with respect to any experiment. More 
refined interpretation of experiments can be achieved by application of the 
present model to calculate other average piT)perties. We anticipate the 
more detailed model will lend support to this approach. 

Publications: 

Ferretti, J. A., Jemigan, R. L., and Weiss, G. H.: Polypeptide 
helix lifetimes in the helix-random coil transition region. Application to 
NNR spectra. Journal of Polymer Science (accepted for publication). 

Jemigan, R. L., Ferretti, J. A., and Weiss, G. H. : Models of 
time dependent behavior of polypeptides in the helix-random coil transition 
region. Application to NMR spectra. Polymer Preprints , 13 , 941-945 (1972) 

Ferretti, J. A., and Jemigan, R. L.: NMR spectra of poly- - 
benzyl -L-glutamate in the helix-random coil transition region. Polymer 
Preprints , 13, 946-950 (1972). 



110 



Serial No. 5.3 



1 . Physical Sciences Laboratory 

2. Not Applicable 

3 . Bethesda 



PHS-NIH 
Individual Project Report 
Jioly 1, 1972 through June 30, 1973 

Project Title: Molecular Mechanics 

Previous Serial Number: 5.3 

Principal Investigator: Robert L. Jemigan, Ph.D. 

Other Investigators: George H. Weiss, Ph.D. 

Cooperating Units: None 

Man Years: 



Total: 


0.4 


Professional : 


0.4 


Other: 


0.0 



Project Description: 

Objectives: 

The purpose is to predict dilute solution behavior o£ macro - 
ncleoiles by means o£ detailed moleciilar models. 1) Equilibrium dimensional 
properties are derivable from intra-chain distance distribution functions 
Cd.f.j. Even mon^nts of distance are calculable for realistic models. These 
"deld approximate d.f. 's wiiich can in turn be applied to obtain bounds to 
other d:Lstan.ce related functions. 2) Non- equilibrium properties can be 
treated in terms of time correlation functions. A general treatment of 
these vrith a realistic molecular model is desired. 

Methods : 

Experimental information on small molecules is combined with a 
one-dijnensional statistical conformational model to permit calculation of 
properties of large linear molecules. Each bond is permitted to assume 
Several fixed -geometry states. For equilibrium properties, nearest neighbor 
interactions between these states, as incorporated by matrix methods, suffice 
to produce agreement between numerous calculated results and experimental 
data. 1) Even positive moments of intra-chain distances have been calculated 
by metttods proposed previously. A general method of expanding the d.f. in 
terms of moments and any approxinate d.f. was developed. Similar express- 
ions appropriate to treat other properties, including light scattering, were 
derived. 



Ill 



2) Non-equilibrium properties are treated by assigning rates for transitions 
between states for each bond. A realistic treatment, including neighbor 
dependence of these rates, was derived; this theory is valid for linear 
HDlecules of any size. 

Major Findings: 

1) Previous expansions of distance d.f.'s were unsatisfactory 
because the approximating d.f . generally used was the Gaussian function 
which is inadequate for shorter chains. The present formulation permits 
one to obtain better approximate d.f.'s by using the freely jointed ciiain 
dof.'s as the approximate functions. 2) the neighbor dependence of 
barriers to internal rotation was calculated for hydrocarbon chains. For 
calculating time correlation fimctions, this neighbor dependence is inportant 
and can be included in a sinple matrix formulation. This method was applied 
to treat the non -equilibrium electric birefringence (Kerr effect) . 

Significance to Biomedical Research: ', 

These calculations allow a prediction of a variety of physical 
properties for linear molecules of any length and vdth any given sequence 
of monomer units, v^ether in biological or synthetic macroinolecules . Per- 
forming such calculations in advance of laboratory experiments can provide 
the exiperimental scientists a guide to indicate the value of particular 
experiments . 

Honors and Awards: None 

Publications : 

Jemigan, R. L.: Electric Birefringence of Flexible Polymers 
in Molecular Electro -Optics , Marcel Dekker , IncT , New York (to appear) . 

Jemigan, R. L. and Weiss, G. H. : An equivalent chain model of 
conf igurational distributions . Polymer Preprints (to appear) . 



112 



Serial No. ^'^ . 

1. Physical Sciences Laboratory 

2. Not Applicable 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Biophysical Analysis 

Previous Serial Number: 5.4 

Principal Investigator: Ralph J, Nossal, Ph.D. 

Other Investigators: George H. Weiss, Ph.D. 

Cooperating Units: Jere Segrest, Ph.D., LPB, NIAMDD 

Nfan Years: 

Total: 0.5 

Professional: 0-5 

Other: 0-0 

Project Description: 

Objectives: 

To simply theoretical foimdations and experimental models for 
various observations v^ich arise in physiology and biophysical chemistry. 
The following problems have received particular attention during the past 
year. 

1) Models for cell growth and motility: The objective is to 
provide mathematical descriptions of properties of biological cell popula- 
tions. 

2) Isolation of ion conducting membrane channels: The objectives 
are to reconstitute activity of conducting biomolecular complexes by incor- 
porating them in lipid model membranes , to understand chemical and physical 
processes underlying cell membrane excitation. 

Methods : 

1. Kinetic equations are derived and solved by enploying 
advanced mathematical methods. 

2. Laboratory techniques of microbiology and electrophysiology 
are combined with techniques of mathematical physics. Procedures have been 



113 



enployed to obtain materials from cell surfaces, and assay them on lipid 
bilayer membranes. 

Major Findings: 

1. Expressions have been derived v^hich relate the movement of 
bands of chemotactic bacteria to relevant metabolic and response parameters, 
Experiments have been performed which substantiate the theoretical predic- 
tions . 

2, A theory has been developed to related migration of individual 
cells to the parameters characterizing motions of large populations conposed 
of many cells. 

Significance to Biomedical Research: 

1. Kinetics of chemotactic bacteria: Chemotaxis seems to be 
inplicated in the recognition of bacteria, by leukocytes, the migration of 
monocytes to regions of tissi:^ damage, and processes of cell aggregation. 
Bacterial chemotaxis is an analog phenomenon which is more amenable to 
analytic study 

2. Techniques for assaying cell surface constituents have 
applications in many studies of membrane transport and excitation. 

Honors and Awards: None 

Publications : 

Nossal, R. J.: The growth and movement of rings of chemotactic 
bacteria. Exp. Cell. Res . 75_, 138-142 C1972). 

Nossal, R. J. and Weiss, G. H. : Analysis of a densitometry 
assay for bacterial chemotaxis. J. Theor. Biol , (in press). 



114 



Serial No. _ 5.5 

1. Physical Sciences Laboratory 

2. Not Applicable 

3. Bethesda 

PHS-NIH 

Individual Project Report 

J-uly 1, 1972 throu^ Jme 30, 1973 

Project Title: Carbon-13 Nuclear Magnetic Resonance Studies o£ Peptides 
and Proteins, Including the Use of SelectiA^ C^^ 
Enrichment 

Pxevious Serial Nianber: 5.5 

Principal Investigator: Jack S. Cohen, Ph.D. 

Other Investigators: I. Chaiken, Ph.D., LCB:NLAMDD; M. Freedoen, University 
of Toronto, Toronto, Canada. 

Cooperating Units: Labcrator)^ of Chemical Biology, NIAMDD 

Man Years: 



Total : 


0.3 


Professional : 


0.3 


Other: 


0.0 



Project Description: 
Objectives: 

1. To apply carbon-l3 Fourier transform NMR methods to macro - 
imlecules . 

2. To obtain information regarding the structure and function 
of peptides and proteins. 

Methods : 

13 

1. Synthesis of peptides containing selected C enriched amino 

acids: two such peptides were prepared from the 1-15 amino terminal 
sequence of ribonuclease contained C^^-Phe at position 8 and C^^^ at position 

12. ^ 

13 

2. Use of C Fourier transform NMR to study these resonances 

in the peptides and the active enzyme complexes. i 

13 

3. The sttjdy of pH and other dependencies of C resonances of 

amino acids and peptides. 



115 



Nfejor Findings: 

1. The C resonances of the enriched amino acid incorporated 
into the peptide and active enzyme complex would be seen above the backgi-oirid 
due to the natural abundance resonances, lids mkes this approach, feiisi.bie 
in studies o£ structure and function of proteins. The effects on C ^ 
chemical shifts and relaxation times could be interpreted in terns of the 
formation of a helical region of the 1-15 peptide on confilex fonrBtion. 

2. Conparison of spectra of the 1 13,1-15, and 1-20 peptides 
with each other and with amino acid spectra allows the definitive assign- 
ments of most of the resonances. Simulation of C^' NMR spectra U'-ing the 
MLAB system greatly assists these assigraients. 

3. Hie pH-dependence of C^-' resonances of several amino acids 
and peptides is somewhat surprising. Little effect of the carbo>:)^l titration 
was observed, but fairly large effects of the amino terminus titration. 
Calculation of electron density has been carried out and mainly are consis- 
tent with the qualitative shifts observed. This is being investigated for 
possible applications. 

4. Extension of these studies to several proteins at natural 
abundance of C^^ indicates definite differences between the spectra of the 
native and denatured forms, but so far it has not been possible to utilize 
these differences for further studies. 

Honors and Awards: None 

Publications : 

Freedman, M. H. , Lyerla, J. R. , Jr., Chaiken, I. M. , and Cohen, 
J. S.: Carbon-13 nuclear magnetic resonance studies of selected amino 
acids, peptides and proteins. Europ. J. Biochem , 32 , 215-230 (1973). 

Chaiken, I. M. , Freedman. M.H. , Lyerla, J. R., Jr., and Cohen, 
J. S.: Preparation and Studies of ^^F-labelled and enriched ^^C-labelled 
semisynthetic ribonuclease-S' analogues. J. Biol. Chem. , 248 884-895 (1973). 



116 



Serial No. 5.6 



1. Physical Sciences Laboratory 

2. Not Applicable 

3. Bathe sda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Excitation and Transport Properties o£ Fluids; Laser 
Scattering 

Previous Serial Nuinber: 5.6 

Principal Investigator: Ralph J. Nossal, Ph.D. 

Other Investigators: Mildred L. McNeel 

Cooperating Units: Leonard Kohn, M.D. , LPB, NLAMDD, S-H. Chen, Ph.D., M.I.T., 

J-P. Boon, Ph.D., Universite Libre de Bruxelles 
Man Years: 



Total: 


0.5 


Professional: 


0.5 


Other: 


0.0 



Project Description: 
Objectives: 

1. To provide basic knowledge concerning the excitation 
properties and transport properties of both sinple fluids and conplex 
solutions containing biological macromolecules . 

2. To develop a laser light scattering spectrometer to 
measure hydrodynamic coefficients of biological macromolecules , the rate 

constants of bimolecular reactions, and the swimming speed distribution 
of motile microorganisms. 

Methods : 

Theoretical, techniques of mathematical physics and statistical 
mechanics are enployed in order to develop new physical theories. Theor- 
etical studies are performed in support of new experiments, particularly 
those involving the scattering of laser light from biological materials. 
A correlation-function laser scattering spectrometer is being developed 
under contract. 



117 



Major Findings: 

Techniques have been developed for losing a laser light scatter- 
ing correlation spectrometer for the purpose of nieasuring the diffusion 
coefficients of biological molecules having a wide range of sizes. The 
theoretical basis of light scattering from motile microorganisms has been 
investigated, and experiments have been performed v/hich establish the 
feasibility of using laser light scattering spectroscopy as a new quantita- 
tive assay for bacterial motility. Precision measurements of bacterial 
motility have been acconplished, and the effects of chemoattractants upon 
motility have been investigated. 

Significance to Biomedical Research: 

Almost all biological phenomena occur in a fluid environment, 
A number of fundamental questions concerning the physical behavior of 
fluids yet remain unanswered. Their elucidation will ultimately facilitate 
better understanding of the functions and properties of biological systems. 

The development of the laser correlation spectrometer shoiold 
enable rapid and precise measurement of various physical parameters which 
characterize systems of biological molecules and motile microorganisms. 

Honors and Awards: None 

Publications : 

Nossal, R. J. and Chen, S.H. : Light scattering by motile bacteria. 
J. de Physique , 33, Cl-171-175 (1972). 

Nossal, R. J. and Chen, S.H. : Laser measurements of chemotactic 
response of bacteria. Optics Comm . _5, 117-121 (1972), 

Nossal, R. J., Boon, J. P., and Chen, S.H: Light scattering spectrum 
due to wiggling motions of bacteria. Biophys. J . (to appear). 

Nossal, R. J. and Chen, S.H. : Effects of chemoattractants on the 
motility of E. Coli Bacteria, Nature (to appear). 



118 



Serial No. 5.7 

1 . Phy s i cal Sciences Laboratory 

2. Not Applicable 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Intermolecular Forces in Biological Structures 

Previous Serial Number: 5,8 

Principal Investigators: V. Adrian Parsegian, Ph.D., George H. Weiss, Ph.D. 

Other Investigators: S. L. Brenner, Ph.D., University of Kentucky, 

Lexington, Kentucky 
Cooperating Ifeiits: None 

Man Years 

Total: 1.9 

Profes sional : 1.5 
Other: 0.4 

Project Description: 

Objectives: 

To identify snd learn to calculate those intermolecular forces 
governing the structure of imtter at the biological level. These have 
been Couloiifcic (electrostatic) interactions between charged species and 
van der Waals (electrodynamic) forces. 

Methods: 

Classical and quantiM-mechanical treatment of electromagnetic 
and statistical -mechanical behavior as well as molecular models of specific 
structural interactions. 

Major Findings: 

Long-range physical forces are strong enou^ to hold biological 
cells together. These forces may be sufficiently specific between cells 
of like tissue type to account for the fundamental observation that cells 
will sort themselves out of a mixture to form aggregates of only one tissue 
type. 



119 



In addition to the use o£ measured absorption spectra to conpute 
electrodynamic forces, we have learned to use data derived from reflection 
spectra to make more reliable estimates. 

The theory of van der Waals interactions has been developed to give 
the interaction between anisotropic and inhoTnogeneous bodies. Also tJiere 
is now a formulation of the interaction betx\reen long thin molecules in 
solution. Ionic salts in. the batjiing medium can exert a strong influence 
on the long-range attractive forces between cells. 

Significance to Biomedica.l Research; 

The physical theory of intercellular forces provides a strong 
logic for understanding and controlling cell -cell interaction. In partiailar 
it is leading to experiments where the outer cell surface may be altered by 
binding of known substances v^dch will in turn promote aggregation of cells. 
We are trying to steer this effort to convert non-aggregative cell popula- 
tions such as metastisizing cancer cells ijito cellular aggregates sticking 
either to themselves or better to an inserted artificial substrate. 

We are beginning to have in hand a set of physical techniques 
suitable for analyzing the organization of constituent molecules into 
cellular organelles and substructures. 

Honors and Awards: None 

Pijblications : 

Parsegian, V. A. and Weiss, G. H.: On van der Waals interactions 
between macroscopic bodies having inliomogeneous dielectric susceptibilities. 
J. Colloid and Interface Sci . 40^, 3S-40 C1972). 

Parsegian, V. A. and Gingell, D.: A physical force model of 
biological membrane interaction, in Recent Advances in Adhesion ed. L. H. 
Lee, Gordon and Breach Pub. Co., New York and London (in press) 1972. 

Parsegian, V. A.: The non-retarded van der Waals interaction 
between thin rods at all angles. J. Chem. Phys . 56, 4393-4399 (1972). 

Parsegian, V. A. and Gingell, D. : Some features of physical 
forces between biological cell membranes. J. Adhesion 4, 283-293 (1972). 

Parsegian, V. A. and Gingell, D. : On the electrostatic inter- 
action across a salt solution between tvro bodies bearing unequal charges. 
Biophysical Journal 12, 1192-1203 (1972). 

Gingell, D. and Parsegian, V. A.: Coii5)utation of van der Waals 
interactions in aqueous systems using reflectivity data. J. Theor. Biol . 
36, 41-51 0-972). 



120 



Parsegian, V. A. and Ninham, B. W. : van der Waals forces in many- 
layered structures: Generalizations of the Lifshitz result for two semi- 
infinite media. J. Theor. Biol . 38^, 101-110 (1973). 

Gingell, D. and Parsegian, V. A.: Prediction of van der Waals inter- 
actions between plastics in water. J. Colloid and Interface Sci. (in press), 
(1973) . 

Parsegian, V. A.: Long-range physical forces in the biological 
milieu. Annual Reviews of Biophysics and Bioengineer'ing (to appear) (1973) . 



121 



Serial No. 5.8 

1 . Physical Sciences Laboratory 

2. Not Applicable 

3. Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Consulting Services 

Previous Serial Number: 5.8 

Principal Investigators: Richard I. Shrager, George H. Weiss, Ph.D. 

Other Investigators: Mildred McNeel, James A. Ferretti, Ph.D., 
Adrian Parsegian, Ph.D. 



Cooperating Units: 



Alan Schechter, Ph.D., Laboratory o£ Chemical Biology, 
NIAMDD, Robert L. Berger, Ph.D., Laboratory of Techni- 
cal Development, NHLI, Paul Berk, M.D. , Metabolism 
Branch, NCI, William Caveness, M.D. , Laboratory of Ex- 
perimental Neurology-, NINDS, David Rodbard, M.D. , 
Reproduction Research Branch, NICHD, Gary Knott, 
Heuristics Laboratory, DCRT, Eugene Fischmann, M.D. , 
Freedman's Hospital, Mario Morini, Ph.D., Northwestern 
University, James McCraig, Ph.D., Paul Smith, M.D., 
Toshio Asakura, M.D,, Johnson Foundation, L&iiversity of 
Pennsylvania, Luciano Forlani, Universita di Roma. 



Man Years: 



Total : 

Professional: 

Other: 

Project Description: 



1.5 
1.5 
0.0 



Ob j ectives : 

To provide consulting services in applied mathematics, biometry, 
theoretical chemistry, theoretical physics, and various aspects of numerical 
analysis to workers primarily in e^qjerimental fields. 

Nfethods : 

A considerable amount of consulting is done in the application 
of curve fitting techniques to the results of biological and biochemical 
experiments. Other consulting involves applications of statistics to 



122 



njEdical data and the application of a program developed by Dr. Ferretti of 
this Laborator/ for tlae determination of physically interesting parameters 
from NMR data. 

Major Findings: 

The kinetics of hepatic -synthesized bilirubin has been stiodied 
and the differences delineated for nomal and diseased human subjects. 

The MLAB s>'stem has been developed in collaboration with Gary 
IQiottj for use on the PDP-10. This system is a descendant of MODEIAIDE 
now available on the ISvl system. The MODELAIDE system is undergoing 
extensive changes. A stiff differential equation solver, useful for 
kinetic eq'xations with greatly differing rate constants, is being developed 
and added to the program, and more error analyses are being provided. 

Long term effects of head injuries during the Korean War have 
been studied. The influence of pre- injury mental status (as measured by 
AFQT scores) in connection with a head injury has been shown to be very 
important as a determinant of work status. 

Honors and Awards: None 

Publications : 

Knott, G., Shrager, R. : On-line modelling by curve -fitting. 
Conputer Graphics 6, 138-151 (1972). 

Jones, E. A., Shrager, R. , Blooner, J. R., Berk, P. D., Howe, 
R. B., Berliln, N.I.: Quantitative studies of the delivery of hepatic 
syntliesized bilirubin to plasma. Journal of Clinical Investigations 51 , 
2450-2458, (1972). 

Weiss, G. H. , Caveness, W. F.: Prognostic factors in the 
persistence of posttraumatic epilepsy. Journal of Neurology 57 , 164-169, 
(1972) . 

Fischmann, E. J., Weiss, G. H. , Barber, M.R. : In-vivo determina- 
tion of orthogonal vector, Proceedings of the XII ' th International 
Colloquium Vectorcardiogi-aphicuin 157-163 (1972) . 

Fischmann, E. J., Weiss, G. H., Downing, J.W., Mehrotra, P.: 
Reader variation in fourteen qi.iantitative VCG abnormality criteria. Pro- 
ceedings of the XII 'th International Colloquium Vectorcardiographicum. 
IM^TTb (1972) " 

Dresser, A, C., Meirowsky, A. M. Weiss, G. H. , McNeel, M. L., 
Simon, G. A., Caveness, W. F,: Gainful employment following head injury; 
prognostic factors . Archives of Neurology (to appear) . 



123 



Rodbard, D., Weiss, G. H.: Mathematical theory o£ immimDradiometric 
(labelled antibodies) assays. Analytical Biochemistiy (to appear). 



124 



Serial Number 5.9 

1. Physical Sciences Laboratory 

2. Not Applicable 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through Jme 30, 1973 

Project Title: Fundamental Studies 

Previous Serial Number: 5.9 

Principal Investigators: George H. Weiss, Ph.D., Richard I. Shrager, 

Andrew G, De Rocco, Ph.D. 

Other Investigators: James E. Kiefer, Mildred McNeel 

Cooperating Units: Milton Sobel, Ph.D., University of Minnesota; 

Menachem Dishon, Ph.D. Weizmann Institute o£ Science, 
Israel 

Man Years: 

Total: 0.4 

Professional: 0.4 

Other: 0.0 

Project Description: 

Ob j ectives : 

This year's effort was mainly devoted to developing the theory 
of clinical trials for the conparative testing of two or more treatments, 
with the object of minimizing the number of patients given the poorer 
treatment . 

A second project was devoted to developing a description of 
isomerizing ions in Brownian motion, in which the diffusional properties 
depend on the state. 

Major Findings: 

We have developed trial designs that are useful when the results 
of an experiment are not immediately Imown, and which bias treatment 
allocation towards the better treatment. So far this work has been for 
trials with a dichotomous response "success" or "failure". We have pre- 
liminary results aA^ilable for the situation in which response is measured 
in terms of survival, as is relevant for cancer chemo-' or radio-therapy. 
These show that it is indeed possible to bias the treatment allocation even 
though responses to treatment are not immediately available. 

125 



Honors and Awards: None 

Publications : 

Weiss, G. H. and Sobel, M. : Some results on using the play-the- 
winner sanpling rule with binomial problems. Proceedings of the Sixth 
Berkeley Symposium , vol. 1, 717-736 (1972). 

Weiss, G. H., Hoel, D. G,, and Sobel, M. : A two stage procedure 
for choosing the better of two binomial popiilations , Bion^trika , 59 , 317-322, 
(1972) . 

Weiss, G. H. and Sobel, M.: Play -the -winner rule and inverse 
saii5)ling for selecting the best of k > 3 binomial populations. Annals of 
Mathematical Statistics , 1808-1826 (1972). 

Weiss, G. H., Hoel, D. G. and Sobel, M.: Comparison of sanpling 
methods for choosing the better of two binomial populations with delayed 
observations . Biometrics (to appear) . 

Weiss, G. H., Simon, R. and Stroot, M. : Numerical inversion of 
Laplace transforms with application to percent labelled mitosis experiments, 
Conputers and Biomedical Research _5, 596-607 (1972) . 

Kiefer, J. E. and Weiss, G. H.: Some asymptotic Bessel function 
ratios. Israel Journal of Mathematics 12, 46-48 (1972). 

Weiss, G. H.: The diffusion constant for two-state brownons. 
Journal of Statistical Physics (to appear) . 

Weiss, G. H. : Some approximate solutions to the discrete master 
equation. Journal of Statistical Physics 6, 179-188 (1972). 

Weiss, G. H. and Blumenfeld, D. E.: Gap stability in the light 
of car -following theory. Transportation Research (to appear) . 



126 



Serial No. 5.10 ^_ 

1 . Physical Sciences Laboratory 
1. Nor Applicable 
3. Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title; Nuclear Magnetic Resonance (NMR) Studies of Proteins 

Previous Serial fiunfeer: 5.30 

Principal Investigator: Jack S. Cohen, Ph.D. 

Other Investigators: A. Schechter, M.D., J. Griffin, Ph.D., D. East, M.D., 
M. Hayes, Ph.D., Laboratory of Cheinical Biology, 
NIAMDD; H. Yah, Ph.D., L. Cohen, Ph.D., Laboratory 
of Chemistry, NIAMDD; M. McNeel, R. I. Shrager 

Cooperating Linits: Laboratory of Chemical Biology, NIAMDD, Laboratory of 
Chemistry, NIAMDD 

Man Years : 

Total: 1-0 

Professional: 0-6 

Other- 0.4 

Project description: 

Ob j actives : 

1. To obtain detailed information on the structure and function 
of er^zyines, both qualitatively in terms of the groiqjs involved and quantita- 
tively in terms of the molecular interactions. 

2. To utilize the histidine NMR titration curves as sensitive 
monitors of protein interactions at the molecular level. It is noted that 
histidine is the only amino acid with a pK around neutrality which gives 
rise to well resolved titrating resonances. 

3. To extend the applicability of high resolution NMR as a 
quantitative analytical tool to proteins and to protein interactions with 
hormones and ligands in solution. 

Methods: 

The region of NMR spectra containing the histidine side-chain 
imidazole C2-H resonances of a number of proteins is subjected to detailed 
analysis. Using the new piilse -Fourier transform accessory 220Nfiz spectra 



127 



are obtained more efficiently than with previous time -averaging techniques, 
and spin-lattice relaxation times can be directly measured. Some spectra 
are digitized and fitted with a series of Lorentzian curves using the 
^DDEIJ\IDE least squares fitting program with the on-line IBM 2250 display 
unit. The chemical shifts of the imidazole resonances as a function of 
pH are fitted to several mechanistic nK)dels using the MLAB system of the 
DEC PDP 10 coirputer. Difference spectra between different states are 
obtained utilizing a conputer program. 

Major Findings: 

1. The histidine ring proton WR titration curves of chemical 
shift versus pH of ribonuclease S (consisting of the non-covalent coii5)lex 
of residues 1-20 and 21-124) and ribonuclease A are almost identical. This 
indicates a close similarity in the electronic environments of the histidine 
residues in these two enzymes contrary to a previous report. In addition 
to the main inflection in these curves due to the imidazole group titration, 
an acid inflection in the pH range 4-5 was also obser^/ed in three of the 
four NF1R- titration curves of ribonuclease S. These inflections arise from 
interactions of the ijiiidazole vvith adjacent carboxyl groups. 

2. Addition of phosphate ion to ribonuclease A results in changes 
in the MMR titration curves of the tv/o imidazole C2 proton resonances, which 
are thought to correspond to the two active site histidine residues. The 
apparent pK values of these two imidazole groups increase with addition 

of phosphate, while the pK values of the acid inflections in these curves 
decrease slightly. These changes produce an increased separation of the 
imidazole and acid inflections and result in a pronounced asymmetry in the 
NMR titration curves. These results clarify several apparent inconsistencies 
in the literature. The pK valtjes determined for the acid inflection do 
not correspond to an ionization of the phosphate group. The effects of 
sulphate ion are similar to those of phosphate. The results si^^port the 
hypothesis that the acid perturbations in these NMR titration curves arise 
from intramolecular interactions of the imidazole groips with nearby carboxyl 
groups. .^■ 

3. Conplex formation between ribonuclease and dinucleotide 
phosphonate inhibitor, UpcA (prepared by G. Jones of Syntex, Inc., Calif.) 
was studied using ^H-NMR at 220MIz and ^^P-I^MR at 40.5MHz. Nine specific 
resonances of the conplex were resolved and characterized: 4 imidazole C2 
protons of His 12, 48, 105, and 119, an im.idazole NH of His 119, the adenyl 
C2 and C8 protons, the uridyl C6 proton, and the phosphorus nucleus of UpcA. 
The NMR results showed that His 12 and 119, as well as the adenyl, uridyl, 
and phosphotyl portions of UpcA are rigid in the conplex and His 119 is 
stacked with adenine. The binding process shows two apparent pK values of 
3.2 and 9.5. In the conplex, the pK values of the active site residues 
His 12 and 119, 6.3 and 6.0, respectively, are essentially unchanged; and 
the phosphonate groi^ remains partially negatively charged. Tliere is no 
evidence for enzyme -induced inhibitor distortion or for a direct histidine - 
phosphonyl interaction. These results inply that in the RNase- substrate 



128 



conplex near pH 7.5 (the pH optimum for RNA cleavage), His 12 may function 
as a base for the attacking 2' hydroxyl, as generally accepted, but His 119 
might not be an acid for the 5" oxygen of the leaving groiq). 

4. The conpact tertiary structure of cytochrome c is lost in the 
apo- cytochrome c, as revealed by a conparison of the proton magnetic 
resonance spectra of the two proteins. A large number of well -resolved 
peaks are observed for cytochrome c in the region of the spectra containing 
aromatic resonances. Only one of these peaks, corresponding to a single 
proton resonance, titrates with change in pH. This is identified as the 
C2 proton resonance of one of the histidine residues, with a pK of 6.4. 

By contrast no such fine structure is seen in the spectra of apo -cytochrome 
c and all three C2 proton resonances of the 3 histidine residues are almost 
equivalent and are observed to titrate with pK values in the range 6.1 - 
6.2, These results indicate that the heme group is required to form the 
biologically active conformation of cytochrome c. 

5. In following the acid transitions of cytochrome c a further sharp, 
single proton resonance appears in the spectra between pH 3.5-3.2. This 
shifts closer to the histidine C2 proton resonance as the pH is lowered, 

and is itself probably a second histidine C2 proton resonance. Both of 
these resonances decrease in area at the e3q)ense of a single broad peak at 
pH < 1. Three peaks in the aliphatic region of the spectra, two of which 
are thought to correspond to methyl groips on the porphyrin ring, are also 
observed to broaden and coalesce at lower pH. However, in this case the 
transition occurs in the pH range 3.7-4.0. All the changes observed are 
reversible. These results indicate that the acid conformational transition 
of cytochroHK c occurs in several distinct stages . 

6. Preliminary studies of human myoglobin (prepared by H. Hagenmaier, 
Tubingen University) have been carried out. This protein contains ten 
histidine residues, and only five titrating C2 proton resonances were 
observed, indicating that only five histidine residues are free to titrate. 
Since horse heart n^oglobin has one more histidine residue than human rnyo- 
globin, and shows six titrating histidine resomances, it should be possible 
to assign the extra titration curve in the horse protein. One titration 
curve in three iryoglobins has a very high pK value (8.05 for sperm whale 
myoglobin and is shifted to high field values. From the effects of ligands 
and consideration of X-ray crystallographic results on the binding of azide 
ion it is possible that this curve represents the "distal" histidine residue 
adjacent to the heme oxygen binding site. Further studies are continuing 
to clarify this point. 

7. A series of model compounds are being studied as a means of 
obtaining data with which to interpret observed effects on imidazole 
residues in proteins. These include a series containing imidazole separated 
from a carboxyl groi^ by one , two and three methylene groips , and which are 
locked in a single conformation (prepared by Dr. Kikugawa in Dr. Louis 
Cohen's laboratory). The results so far indicate that there is no signi- 
ficant increase in pK value of the imidazole due to the adjacent carboxyl 



129 



grox4>, and that the degree o£ shielding or deshielding presumably Spends 
on the angular orientation o£ the two groups. 

8. A series of fluorine -containing conpounds are also under study. 
The ^^F as well as the proton ^JMR titration curves have been measured for 
2- and 4- fluoro- imidazole and histidine (prepared by Dr. Kirk in Dr.- Louis 
Cohen's laboratory). These show some very interesting results in that the 
2- and 4-F conpounds give opposite chemical changes on increasing pH. 
Electron density calculations are being made in an attenpt to understand 
these effects. 

9. A previous study of the binding of the peptide hormone oxytocin 
to the "carrier" protein neurophysin, has been extended to include irieasure- 
ment of relaxation times, and expanded to another hormone lysine vasoprie- 
sin (LVP) (materials si^plied by P. Cohen and R. Alazard, CEN, Saclay, 
France) . The results for oxytocin using the pulse-FT method were very 
consistent with those obtained previously in CW mode. Linear relationships 
of the spin-spin relaxation tine of the protons of the tyrosine residue of 
both oxytocin and LVP with molar ration (neurophysin hormone) were obtained, 
and extrapolation to (1:1) stochometry indicates that the tyrosine residue 
is rigidly bound in these conplexes. This depends upon a fast exchange 
process being operative, and measurements at several tenperatures for both 
oxytocin and LVP indicate that this is the case. These results for LVP 
give the opposite conclusion from those of another group working on this 
problem. 

10. Studies of the interactions of seAreral rare-earth ions with the 
Ca (II) -requiring enzyme Staphylococcal nuclease have been made using proton 
nagnetic resonance' spectroscopy. Experiments with La(III) served as a 
diamagnetic control for the effects of the paramagnetic ions studied. I^asure- 
ment of line -widths of the four histidine C2 proton resonances in the presence 
of increasing quantities of Gd(III) or Dy(III) has yielded ratios of metal - 
proton distances comparable to those determined by X-ray crystallography. 
Values of binding constants for Eu(III) and Dy(III) were determined from 
measurements of chemical shift changes of the histidine C2 proton resonances 
at pH 5.25. Inactivation of the enzyme in the lisual DMA assay occurred with 
these rare -earth ions and inhibition constants of 1-2 M at pH 7.0 were 
determined by enzyme kinetic analysis. These studies help to characterize 
the ion-binding site of nuclease in solution. 

Significance to Biomedical Research 

1. A general and quantitative method has been developed for the 
analysis of resolved titrating resonances in protein MMR spectra. Applica- 
tion of this method has shed light on the mode of action of several enzymes 
(S . nuclease , carbonic anhydrase , rtbonuclease) . 



130 



2. An ajialysis of the Nf-1R titration data for ribonuclease with 
several inhibitors has provided a detailed analysis of its mode of cleavage 
of ribophosphodiesten bonds. 

3. A study of several resolved resonances during the acid denatura- 
tion of cytochrome £ provides further evidence for a multiple step, as 
opposed to a one step, equilibrium during denaturation. This is consistent 
with previous results for nuclease. It was also found for cytochrome c 
that the prosthetic groi^ is required to generate the active biological 
conformation of the protein. 

4. Results for human m>^oglobin indicate a possible means to monitor 
the process of oxygen binding. 

Honors arid Awards: None 

Publications : 

Cohen, P., Griffin, J. H., Gamier, M, , Gaizergues, M. , Fromageot, 
P. and Gohen, J, S.: Hormonal interactions at the molecular level: High 
resoltuion P>1R study of bovine neurophysins and their interactions with 
ox>-tocin. FEES lett ers., 25, 282-286 (1972). 

Cohenj J.S., Hagenmaier, H-P., Pollard, H. and Schechter, A. N.: 
Proton magentic resonance study of the histidine residues of sperm whale 
and horse fflj'-oglobins . J. Mo'l. Biol ., 71, 513-519 (1972). 

Schechter, A. N., Sachs, D. H., Heller, S. R. , Shrager, R. I., and 
Cohens J. S. ; NMR titration cur/es of histidine ring protons, III. Ribo- 
nuclease. J. Nfol. Biol ., 71, 39-48 (1972). 

Cohen, J. S., Griffin, J. and Schechter, A. N.: NMR titration curves 
of histidine ring protons. IV. The effects of phosphate and sulphate on 
ribonuclease . J. Biol. Chem . (to appear) . 

Griffin, J. H., Cohen, J. S. and Schechter, A. N. : Nuclear magnetic 
resonance studies of a ribonuclease -dinucleotide phosphonate conplex and 
their inplicatiorts in the mecJianism of the enzyme . Ann. N. Y. Acad. Sci . , 
(to appear) . 

Griffin, J. H., Gohen, J. S., Cohen, P., and Gamier, M.: Proton 
magnetic resonance studies of hormonal interactions at the molecular level: 
Conplex foniHtion between bovine neurophysins and oxytocin. J. Pharm. Sci . , 
(to appear) . 



131 



Serial No. 5.11 

1 . Physical Sciences Laborator>' 

2. Not Applicable 

3 . Bethesda 

PHS-NIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: High-Resolution Carbon- 13 Fourier Transform M^IR 
Spectroscopy at 55 Miz 

Previous Serial Number: 5.11 

Principal Investigator: James A. Ferretti 

Other Investigators: Edwin D. Becker, Ph.D., Laboratory of Chemical 

Physics, NIAMDD; Victor Colbam and Luther Barden, 
Conputer Systems Laboratory, DCRT; Thomas Clem, 
Bioelectrical Engeineering, DRS 

Cooperating Units: Laboratory of Chemical Physics, NIAMDD, Bioelectrical 
Engineering, DRS, Computer Systems Laboratory, DCRT 

Man Years : 

Total: 0.4 

Professional: 0-4 

Other: 0-0 

Project Description: 

Objectives: 

To obtain natural abundance carbon- 13 N^4R spectra at 55 Miz 
(ca. 51 kgauss) xosing the Fourier transform and cross correlation techniques, 

Methods: 

A -working 55 I^z pulse NMR spectrometer has been constructed 
and programs have been written for a Raytheon 704 conputer which is inter- 
faced to the spectrometer. The computer is used to time average and 
Fourier transform the response to a short pulse or to conpute the cross 
correlation function. 

Major Findings: 

Spin lattice relxation times on selected small molecules are 
shorter at 55 MHz which indicates that chemical shift anisotropy is an 
inportant relaxation mechanism. 



132 



Ffonors and Awards: None 
Publications : None 



133 



July 1, 1972 through Jun" 30, 1973 

PUCLiC HEALTH SERVICE - NATiONAL IflSTITUTES OF HEALTH 
Dl VISION OF COfTbTER RESEARCH AND TECI'MOLOGY 

1. PORT 6 

Serial fJunber 

lEUPJSTICS LABORATORY 3. Janes R. Slagle 

Chief 



The Heuristics Laboratory has greatly expanded its role 
as a groLsp which collaborates with NIH researchers outside 
of DCRT, with one major service facility introduced and 
anot^<3r greatly extended. At the sane tine, tite laboratory 
has continued to conduct frontier research in the 
application of conputers to biomedical problems. 

The mass-spectral search program (DCr'T 5.9) ""has been 
widely adopted by scientists in NCI,fJMLI, EPA, and F^A, and 
has won the distinction of being adopted for commercial use 
by the General Electric time-sharing network. As soon as 
the legal problems are ironed out, this program will be 
available to users throughout the U.S., Canada, Japan, and 
Viestern Europe, Sn essence, the program assists the user in 
identifying an unknown chemical substance. The chemical Is 
placed in an instrument called a mass spectrometer, which 
breaks the molecules Into fragments and measures ttie 
mass/charge ratio for the most frequently detected 
fragments- The output of the mass spectrometer is a list of 
peaks. The user selects a few peaks, such as (55,253,293) 
and types them into the computer, using a typewriter-like 
terminal and an ordinary telephone line for communication. 
The computer searches its file of 8782 chemicals and types 
out a list of a few substances which match the spectrum of 
the unknov^n substance. The user can then further restrict 
the number of possibilities. The program is also capable of 
searches based on molecular weight and other information. 

MLAB (OCRT b.h) has won broad acceptance by the Ni'l 
community, with over 1000 interactive sessions recordeJ. 
MLAB has been used by some hundreds of N!H scientists, and 
more than a dozen publications have cited f^LAB in their 
bl bliograph les- , MLAB is an interactive system for 
mathematical modeling. The heart of the system is a curve 
fitting program v^hich will adjust the parameters of a model 
function to minimize the sum of the squared errors. A 
repertoire of .mathematical operators and functions, a 
collection of routines for teletype and cathode ray tube 
plotting, and rrechanisms for savin??; data between sessions 

135 



provide a powerful and convpnient ervironmrnt for data 

manipulation, arithmetic calculations, and f^^r building and 

testing models. A major advantage of f'LAB is that it Is 
easy to use by a non-programmer. 

Other biomedical projects include researcfi in automatic 
pattern recognition, including a technioue for classifying 
drugs as sedatives or t ranau i 1 i ze rs on thp basis of their 
mass spectra; computing aids for automatic tissue typing, 
with the objective of increasing the, survival tine of 
transplants; automatic question-answering and problem 
solving; a microbiology data bank; research in information 
storage and retrieval; a computer model of the human 
menstrual cycle; automatic radiation treatment planning; 
heuristic methods for chemical structure searching; energy 
state calculations; and a nuclear magnetic resonance search 
program v;hich, like the mass-spectral search program, aids 
in classifying an unknown compound. 

Curing the reporting period, the nine members of the 
laboratory have published a total of 21 papers and journal 
articles. Honors and Awards: Dr. Stephen P. Heller v/as 
elected to the Advisory Board of the Journal of Chemical 
Documentation. Dr. Richard Lee was selected Treasurer o^-' 
the ACM Special Interest Group on Artificial Intelligence. 



136 



Serial Mo. DCPT 6.1 

1. Heuristics Laboratory 

2. 

3. Bethesda 

P!fS-fl!M 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: Automatic Quest i on-Ansv/er i nc: and Problem Solvinp; 

Previous Serial Number: 6.1 

Principal Investigator: James R. Slagle, Ph.D. 

Other Investigators: Chin-Liang Chang, Ph.D., John K. Jixon, 

Ph.D., Richard C.T. Lee, Ph.D., Lev/is ^^ 
Norton, Ph.D. 



Cooperating Units: 


None 


flan Years: 




Total : 


1.2 


Profess ional : 


1.2 


Other: 


None 



Project Description: 

Object i ves : 

To develop computerized methods of logical inference 
which can act as a basis for advanced question 
answering systems. 

flethods Employed: 

Theoretical studies in symbolic lop:ic. 

Major Findings: 

The standard technique of theorem proving by heuristic 
search has been found to be deficient because of the 
"combinational explosion" of ways of proving the 
theorem. A technique has been developed for building 
in certain very general concepts such as commutat i vi ty 
and associativity, so that the theorem prover "already 
knows" these ideas. One finds that the effort needed 
to prove a theorem is thus greatly reduced. 



137 



significance to Biomedical Research: 

The techniques developed in this project ere expected 
to find eventual application in prob 1 c-^-sol v i ng 
systens for bicnedical research. 

Proposed course: 

The resources allocated to this project are being 
reduced in order to. increase the effort applied to 
more immediate biomedical projects. 

Publ i ca t i ons : 

Chang, C.L., R.C.T. Lee, and J. Dixon, "The 
specialization of programs by theorem proving." To 
appear in SIAM Journal on Computing. 

Chang, C.L., "The Decomposition Princinle for Theorem 
Proving Systems," Proc. Tenth Allerton Conference on 
System Theory, University of Illinois, Oct., 1972. 

Chang, C. L., and R. C. T. Lee, Symbolic Logic and 
Mechanical Theorem Proving, Academic Press 1Q73. 

Chang, C.L., "Theorem Proving v-.'ith Variable- 
Constrained Resolution," Information Sciences, 
Vol . k, 217-231 (1972). 

Dixon, J.K., "Z-Pesol ut ion : Theorem Proving with 
Compiled Axioms," J.ACf^ pp. 127-1^7, Jan. 1973. 

Lee, R.C.T. , and C.L. Chang, "On the Automatic 
Generation of Hypotheses," Prcc. Tenth Allerton 
Conference on System Theory, University of Illinois, 
Oct, 1972. 

Slagle, J.R., and L. M. Norton, "Experiments with an 
Automatic Theorem Prcver Having Partial Ordering 
Inference Rules," to appear In Comm, ACf^ 

Slagle, J. R., "Automatic Theorem- Prcv i ng for Theories 
with SImplifiers, Commuta t i v i ty, and Associativity," 
to appear in J.AC^'. 

Slagle, J. R., "An Approach for Finding C-Linear 
Complete Inference Systems," J.ACf, July, 1972. 

Slagle, J, R., "Artificial Intelligence," Japanese 
translation published; German translation in press. 



138 



Serial No. DCRT 6.2 

1. Heuristics Laboratory 

3. Bethesda 



PHS-Nm 

Individual Project Report 

Uuty 1, 1972 through June 30,. 1973 



Project Title: Storage and Retrieval Research 

Previous Serial Number: Same 

Principal Jnvest igator : Gary D. Knott 

Other Investigators: None 

Cooperating Units: None 

Man Years: 

Total: .5 
Profess ional : ,5 
Other: None 

Project Description: 

Objectives: 

The development and analysis of various storage and 
retrieval algorithms. 

Methods Employed: 

Mathematical investigation, computer experiments and 
contemplation are used to study various aspects of 
certain storage and retrieval schemes. 

Major Findings: 

Some exact analyses have been done and several 
algorithms have been developed. These are hashing 
algorithms as described below. 



Given a collection of items presented in arbitrary 
order^ we wish to store these items and upon demand 
retrieve those items whose key-values match given 
key-values. A key-value is the value under which I terns 
are filed. Moreover vie may wish to delete 
previously-stored items and include further new Items 
from time to time. 

139 , 



One particular approach tc the strrase ann retrieval 
problem is the hash table Tiethcd of storage and 
retrieval. The spirit of such schemes is to use the 
key value of an item to compute ar address for tie 
storage or retrieval of that iter. Key-values for 
which the same address is computed are caller syncnyns. 
The possibility of collisions due to the occurrence of 
synonymous key-values is a najcr difficulty v/hlch can 
be overcome In various v/ays. 

Because the address calculation Is gene.rally a 
randomizing scrambling of the given key-value, the term 
hashing has becor.e the name of this computation. The 
storage area used to store items is kncv;n as a hash 
table and the various algorithms for accessing hash 
tables and resolving collisions are knov/n as hash table 
storage and retrieval algorithms (or just hash storage 
a 1 gor I thms ) . 

There are tv^o major formulations of hash table storage 
and retrieval algorithms, differing In the m.anner in 
which collisions are resolved. One of these approaches 
is to establish a hash table for the storage of the 
items and tc resolve collisions by somehow finding an 
unoccupied space for those I terns whose natural home 
location (according to the hashing function being used) 
is already full. In such a way that the item can be 
later retrieved. 

Algorithms which use such sche'^ies are called open 
addressing algorithms. The other approach finesses the 
problem of collisions by usint. Indirect addressing 
(i.e., sim.ple list-processing techniques) to allow all 
Items vyhlch collide to maintain a claim tc their home 
location. These methods of handling collisions are 
commonly called chaining methods. 

This work has been concerned with both of the 
above-described methods of collision resolution. 

Significance to Biomedical Research: 

These results can help programmers do a better job In 
building storage and retrieval systor^s (but only If 
they know about them!) and strrrge and retrieval 
systems can be of use to biomedical research. 

Proposed Course: 

To continue to work in this area, studying other 
al gor i thms . 



140 



Serial Mo. nCTT 6.5 

1. Heuristics Laboratory 

2. 

3. Bethesda 



PHS-NIH 
Individual Project Peport 
July 1, 1972 through June 50, 1973 

Project Title: Automatic Pattern Recognition 

Previous Serial Number: Same 

Principal Investigator: Janes R. Slagle, Ph.D. 

Other Investigators: C. L. Chang, Ph.D.^HL, A. ^^ Guarino, 

Ph.D., NCI,S. P. Heller, Ph.D., T. 
L. Jones, Ph.D., P. C. T. Lee, 
Ph.D., HL, G.W.A. fMlne, Ph.D., NHLI f. 
Shaniro,CSL, K.L.H. Ting, Ph., CSL 

Cooperating Units: Computer Systems Laboratory, 

NCI Laboratory for Toxicology 
NHLl Laboratory of Chemistry 

flan Years: 

Total: 2.5 
Professional : 2.5 
Other: None 

Project Description: 

Object Ives : 

To advance the state of the art of those areas of 
pattern recognition which holds promise for biomedical 
applications, such as cluster analysis. 

(lethods Employed: 

The development and programming of new alsorlthms tr 

classify patterns and aid in cluster analysis. The 

application of artificial intelligence techniques to 
pattern recognition. 

Major Findings: 

An algorithm for finding a piecev/ise linear discriminant 

function v^as developed. This program can be used for 

unimodal as well as multimodal pattern recognition 
probl ems . 

141 



An algorithm for fitting a plecevvlse linear function to 
multidimensional data was developed. 

A program was v;ritten to project classes of pattern onto 
their Fisher directions. 

Clustering programs were \;ritten usinr; the concepts of 
minimal spanning path and minimizing the sun of 
intragroup distance. 

A one-dimensional clustering technique was developed in 
vjhich a list X of real number is separated into clusters 
v/here a cluster is any maximal sublist vi\:ose mean 
density is larger than the average density over the 
range of X. The method was extended to the 
n-d imens i ona 1 clustering problem. 

The clustering programs were successfully tested on a 
variety of sets of data, including handwritten 
characters. United Nations voting records, and 
classifying drugs as sedatives or t ranqu I i i zers . The 
properties of multilayored linear machines as pattern 
classifiers Vy/ere studied. Programs were written to 
check the relative power of specific arrangements of 
such machines. 

A program was written to measure lesion sizes using 
radioisotope scintigrams. 



Significance to Biomedical Research: 

Almost every area of biomediclne has imprecise drawing 
of conclusions from masses of data. Automatic pattern 
recognition holds great promise in simplifying and 
speeding this v/ork. 

Proposed Course: 

An intensive effort vn 1 1 be untaken to uncover more 
fields which are ripe for any of our techniques of 
cl uster anal ys i s . 

Publ icat i ons : 

Chang, C.L., "Pattern Recognition by Piecewlse Linear 
Discriminant Functions," to apoear in IEEE Trans. on 
Computers . 



142 



Chang, C.L., and R.C.T. Lee, "A Heuristic Relaxation 

flethod for Non-Linear Mapping in Cluster Analysis," IEEE 

Trans. on Systems, Man, and Cybernetics, pp. 197-200, 

March, 1973 

Modes, L., "Solving Problems by Formula Manipulation in 
Logic and Linear Inequalities," Artificial intelligence. 
Fall, 1972. 

Lee, R.C.T. , "An Algorithm to Generate Prime Implicants 
and its Application to the Selection Problem," 
Information Sciences, Vol. k, 1972, pp. 251-259. 

Ting, KoL.H., Lee, R.C.T., Guarino, A. M,, Milne, 
G.W.A., and Shapiro, M., "Applications of Artificial 
Intelligence, Relationships between flass Spectra and 
l^armacological Activity of Drugs," Science, April 27, 
1973. 



143 



Serial No, OCPT 5.1+ 

1. Heuristics Laboratory 

2. 

3. Bethesda 

PHS-fJIH 

Individual Project Report 

July 1, 1972 through June 30, 1973 

Project Title: MLAB 

Previous Serial Number: 5.5 

Principal Investigator: Gary D. Knott 

Other Investigators: Douglas Reece 

Cooperating Units: 'lone 

Man Years: 

Total: .5 
Profess ional : . 5 
Other: None 

Project Description: 

Object i ves : 

The development and promotion of an easy-to-use on-line 
model ing system. 

Methods Employed: 

The programming language, SAIL has been used to code 
basic algorithms and a shell of elaborate human 
Interface routines which all run on the PDP-10 
time-sharing system. 

Major Findings: 

MLAB enjoys great success at NIH and Is used by many 
investigators in the course of their ov.-n research. 
Being on-line is extremely important. So the fact that 
the system is easy to use and requires no progranning 
on the part of the researcher. 



144 



significance to Biomedical Research; 

The system has been used by chemists studying 
equilibrium phenomena and bond interactivity in 
chemical compounds. Physicians have been employint^ it 
for studying absorption of radioactive tracers, neural 
transfer mechanisms and drug interactions. 

Proposed Course: 

To continue to maintain. Improve and promote the 
system^ 



145 



Serial Uo . OCRT 6.5 

1. Heuristics Laboratory 
2. 

3. Bethesda 
PHS-f'I'f 
Individual Project Peport 
July 1, 1972 throuf^h June 30, 1^73 

Project Title: I nf or; la t i on Storage and Petrieval v/I th Index 

Previous Serial ^Junber: 6.80 

Principal Investigator: Leu'is i\ Norton, Ph. P. 

Other Investigators: Pobert Maanuson, Df'B, PCPT. 

Cooperating Units: Pf'C^nCPT 

I'an Yea rs : ' 

Total : .2 
Profess i ona 1 : .2 
Other: 

Project Description 

Obj ect i ves : 

To extend the usefulness of DCr'T software systens for 
information storage and retrieval, by providinr; a 
retrieval capability based on an index of textual 
nater i a 1 in files. 

Methods Employed: 

The RflAG language is used to create user-specific 
software for Indexing a file and auerying tlic index 
using Boolean combinations of keyv.-ords. Keyv/ord to be 
indexed may be specified explicitly by the user. 
Alternatively, the user nay specify a list of v/ords 
not to be indexed, and all vyords in the text but these 
v;ill be indexed. Sophisticated OS data nanagencnt 
techniques are employed. The routines are interfaced 
with existing nflB software for f i 1 f • creation and 
ma i ntenance . 

Major Findings: 

iJone 



146 



Significance to Bionedicine: 

This system provides a fast, flexible means of 
searching files containing data in textual form which 
can be used in a variety of applications. 

Proposed Course: 

The system is basically complete. fllnor modifications 
may be made to accommodate specific apol i cat ions . 



147 



Serial No. PCRT 5.6 

1. Ifeuristics Laboratory 

2 . 

3. Be t he s da 

Individual Project Reoort 
July 1, 1972 through June 30, 1973 

Project Title: Microbiology Data Bank 

Previous Serial Number: 6.81 

Principal Investigators: Lewis fl. Norton, Ph.D., 

Micah !. Krichevsky, Ph.D. MIDP,. 

Other investigators: None 

Cooperating Units: Env i ronnenta 1 h^echanisns Section, MIDP. 

Man Years 

Total: .h 

Profess ional : . k 
Other: 

Project Description: 

Object i ves : 

To develop softv;are enablins:; stora'^e and on-line 
retrieval of data about bacteria and oth-^r 
mi c roorgan i sms . 

Methods Employed: 

Systems analysis and software development has attempte! 
to meet the unique requirements of tlio international 
community of microbiologists, in orJer that the maxinium 
amount of data about microor;-;an i sns may be accumulated 
and made available. The kinds of data as we 1 I as tlioir 
uses are extremely diverse. Flexible, interactive use 
of the system has been emphasized. 

Major Findings: 

We have encoded data on antibiotic sensitivities of over 
17000 strains of pathon:enic bacteria from N I fi * s Clinical 
Pathology Laboratory, CC. A by-product was t^ sireatly 
decrease the error rate of the input file from t'lO 
CI inical /Pathol ogy Lab. Smaller volumes of data frr^n 

148 



other sources in the United States have been encoded. 
Groups from Australia, England, Japan, Czechoslovakia 
and many other countries, as v;ell as fron nany points in 
the United States, have indicated interest in 
contributing to and making use of such a data bank. 

Significance to Biomedical Research: 

Data about bacteria and other mi croor^^an i sms is of 
importance to many medical specialities. 

Proposed Course: 

Increased use of the system is expected to reveal 
desirable modifications and extensions. The coding 
method will continue to be expanded to allow for 
additional types of data. Use of the data bank will 
have to be possible from different types of computers. 
Increasingly sophisticated retrieval methods for 
utilizing the data will be developed. 



149 



Serial Mo. DCPT 6.7 

1. fleuristics Laboratory 

2. 

3. Bethesda 

P!'3-NIH 

Indivi-iual Project Report 

July 1, 1972 throu-h June 30, 1Q73 

Project Title: Tissue Typing Computer Aids 

Previous Serial Number: None 

Principal Investigators: Le\;is M. Norton, Ph.D., 

John K. rixon, Ph.D., 
Donald E. Kayhoe, f^D., NMAID 

Other Investigators: John G. Ray, Jr., Ph.D., NIAID 

Cooperating Units: Transplantation and Innunolc^y Branch, 
NIAID 

tlan Years: 



Total : 


0. 


,7 


Profess i ona 1 : 


0. 


,7 


Other: 


0. 


.0 



Project Description: 

Obj ect i ves : 

To improve knowledge of tissue typin^^. Ultimately this 
knowledge may contribute toward making or<;an transplants 
more successful . 

Methods Employed: 

Heuristic programming techniques enabled the development 
of an economical, accurate program for automatic 
evaluation of tissue typing results. The program is 
independent of current knowledge about tissue 
(histocompatibility) antigens; i.e., such knov.'ledge is 
input to the program, along with specific typing 
results. Additional programs were written to further 
analyze data in terms of conditional probability and 
cl uster i ng. 



150 



Major Findings: 

The tissue typing program attained a satisfactory level 
of performance, and is more consistent than humans in 
its evaluations. It will be possible to process large 
amounts of typing data for various statistical studies. 
Conditional probability methods identified discrepancies 
between theory and data. 

Significance to Blomedicine: 

There Is controversy among Immunol ogi sts as to the 
Importance of matched tissue types between donor and 
recipient of organ transplants. This is due partly to 
shortcomings of biomedical knowledge about 
histocompatibility antigens, and partly to a lack of 
vyel 1 -des I gned/ large-scale experiments for evaluating 
the role and importance of these antigens for transplant 
survival or rejection. Large amounts of raw data exist, 
and should be carefully evaluated. Such evaluation can 
also point to deficiencies and possible improvements in 
the identification of particular antigens. 

Proposed Course: 

Additional heuristic programs will be developed to 
correlate tissue typing records vnth transplant survival 
data being collected for NIAID. 



151 



Serial 'lo. PC^T 6.8 

1. Heuristics Laboratory 

3. Pethes'ia 

InHividual Project ^ec^rt 
July 1, 1972 throu;h June 3i^, 1073 

Project Title: Nuclear faf^netic Peson--5nro-nr3n'n i cs M"r^ 
Ana lysis Prr'^rar^ 

Previous Serial Number: 6.90 
FVincipal Investigator: Stephen R, t'eller, Ph. P. 

Other Investigators: Richard J. Feldnann, CC^; 

Arthur E. Jacobson, Ph.D., LC, mAfP, 
Robert J. Hi.i;hct, Ph. P., LC, 'l^'il ; 
Edward A. Sol i-.ol ov.s'-. i , LC, rJiai 

Cooperating Units: CCB; Laboratory of Chenistry, :,'ISf^P; 
Laboratory of Che'^^istry, fJI'Ll 

Man Years 

Total : .8 

Profess iona 1 : . 8 
Other: 

Project Description: 

Object i ves : 

To interactively analyzn nuclear .'lairnetic resonance 
(h'MR) spectra for the chenical shifts and coupling 
constants contained in the spectra. 

flethods Employed: 

An interactive Nf'R analysis pro'^ran has been written 
for the PDP-10 which allov.s for tbo graohical display 
of calculated NfT, snectra. The prof^ran also contains 
provisions for two types of har"" cony Calcorp output 
vvhich are useful for pernanent records and 
pub 1 I cat I ens . 



152 



Major Findings: 

Analysis of NflP spectra of ccnplex spin systens are 
readily handled In a graphical intoractive node. The 
overall elapsed tine for analysis has beon reduced 
from weeks to days using conputer graphics. 

Significance to Blonedlcal Research: 

This program Is believed to be the most sophisticated 
and comprehensive Nf'lP analysis program available. The 
combination of computer graphics and hard copy 
plotting options Is unique. 

Proposed Course; 

As the program Is used more. It is anticipated that 
various users will request additional options be 
added, such as simultaneous display of exnerlnental 
and calculated spectra. 

Publ icat I ons : 

1. Heller & Jacobson, Anal.-Chem. kh, paee 2219 
(November 1972). 

2. Heller, DCPT/CIS Graohical Interactive Mf'P 
Analysis Users Manual, March 1973. 



153 



Serial flo. DCPT G.D 

1. Heuristics Laboratory 

2. 

3. Be t he s da 

PHS-MI!I 
Individual Project Penort 
July 1, 1G72 through -June 30, 1973 

ft-cject Title: llass Spect ror.et ry Data Interpretation 

Previous Serial Hunber: G.91 

FV-incipal Investigator: Stephen R. Heller, Ph . T . 

Other Investigators: Pichard J. Feldnann, CCP 

Henry P.. Fales, Ph . P . , G.V.'.A. f'ilne, Ph . P . , 
David C. Peddcr, Ph. P., LC,N"L! 

Laboratory o-*" Chenistry, '^!''LI 



ope rat i ng Uni ts : 


Labo ra 


lian Years 




Total : 


1.2 


Profess i ona 1 


: 1.2 


Other: 


Hone 



Project Description: 
Object i ves : 

The developnent of an interactive conversational 
search and retrevial systen for nass soectronetry data 
for the purpose of structural elucidation. 

flethods Employed: •'' 

A highly conversational, interactive search systen for 
a large file of nass spectral data v/as v;ritten for the 
PDP-10. The system iias fifteen onticns. They are: 

1. Peak and Intensity Search 

2. Molecular I7eight Search 

3. Complete and Partial Molecular Formula Search 
U. Peak and Miolecular Veisiht Search 

5. Peak and Molecular Formula Search 

6. Molecular I7eight and Formula Search 

7. Dissimilarity Index Comparision 

8. Spectrum Printout 

9. Display of Spectra 
10. Plotting of Spectra 



154 



11. Microfiche Retrieval 
12 . CRAB-Comments and Complaints 
IS.HARVEST-Entering of New Data 
11*.NEWS-News of the System 
15.MSCE Code List 

Major Findings: 

It is possible to instantaneously search 8782 mass 
spectral file with resulting ansv^ers identifying the 
molecule or suggesting possible solutions. The search 
system is accessed from any computer terminal in the 
Heart, Cancer and Arthritis Institutes. 

At present the system has been used by over 200 
scientists in the United States and Canada with 
average daily usage of 25 sessions. 

Significance to Biomedical Research: 

The ability to have a large library of mass spectra 
always available vnth a computer terminal and 
time-sharing PDP-10 computer has enabled scientists to 
readily identify and assist in the structural 
elucidation of biologically significant molecules. 
The system has greatly aided the scientists involved 
in the mass spectral analysis of blood samples from 
drug overdose patients from local hospitals. 

Proposed Course: 

It is planned to expand the size of the file with the 
emphasis on adding drugs, pesticides and related 
chemicals of biological interest. Also, structural 
informiation in the form of Chemical Abstract Registry 
Numbers, will be added to the file so that the system 
search options can be expanded. 

The system is also being transferfed to the 
international GE timesharing network for use from 
Japan, U.S., Canada and Western Europe via local 
telephone connections. This is being done so that a 
vnder use of this succesful project can be found and 
so that NlH's role as a research and not production 
organization can be maintained. 



155 



Publ icat Ions : 

1. Heller, S.R. "Conversational f^ass Spectral 
Retrieval System and Its Use as an AM in Structure 
Determination/' Anal. Chen., tjU, 1D51-19G1 (1972). 

2. I^eller, S.R., Fales, H.r\ and fMlne, G.Vi.A., "An 
Interactive Mass Spectral Search System," J. Chen. 
Ed., i+9, 725 (1972). 

3. Heller, S.R., Fales, H..":. and f^lne, G.W.A., "A 
Conversational Mass Spectral Search and Retrieval 
System, M. Combined Search Options," Org. Mass. 
Spec, 7, 107-llU (1972). 

h. Meller, S.R. "DCRT/CIS Mass Spectral Searcli System 
User's Manual," November 1972 DCPT, rilH, Eethesda, 
Ma ryl and. 



156 



Serial rio. DCTT 6.10 

1. F'euristics Laboratory 

3. Bethesda 



pi!S-riiii 

individual Project Peport 
July 1, 1972 through June 30, '1973 

Project Title: Energy State Calculations 

Previous Serial Number: 6.92 

Principal Investigator: Stephen R. Heller, Ph.D. 

Other Investigators: Richard J. Feldmann, CCR; Arthur E. 

Jacobson, Ph.D., LC,rJIAfTn; Robert Katz, 
Ph.D., LC,rilAf;PP; IJlrich Weiss, 

LPB,rnAr'n 

Cooperating Units: Physical Science Laboratory; Conputrr 

Center Branch; Laboratory of Chenistry, 
NIAMD; Laboratory of Physical Chenistry, 
NIAfiD 

I!an Years: 

Total: 1.5 

Profess ional : 1.5 
Other: 

Project Description: 

Objectives: 

The development and use of conputor protT;rai":s for 
calculating the conformation and energy of a nolcculo. 

flethods Employed: 

A number of computer prn-raris to do enern;y 
minimization and enerf^y state calculations has been 
investigated. The CMDO/lfJDO enerp;y state calculation 
program has been redesigned and re\/rlttcn fr^r the 
PDP-10 and is to be used in conjunct i-^n v.ith X-Ray 
flodeling System on the PPP-IO. 



157 



t'.ajor Findings: 

There are no computer pro^-rans available for energy 
nininization of systems of bio1o2:ical interest. If a 
conformation is kno\m, tlie CNDO/IMDO progran is a 
reasonable propran tc use for the calculation of that 
conformation. 

Proposed Course: 

To undertake a further stuiJy of b i ol or; i ca 1 1 y 
interesting molecules such as catecholamines anA 
cycl ohexad i ene derivatives. 



158 



Serial No. DCPT C.ll 

1. fieuristics Laboratory 

2. 

3. EethesHa 



PMS-NIl! 
Incilviciual Project Report 
July 1, 1972 throuf^h June 30, 1973 



Project Title: Turnaround Reduction 

Previous Serial Number: None 

Principal Investigator: Thonas L. Jones, Ph.l^. 

Other Investigators: None 

Cooperating Units: None 

flan Years : 

Total: .3 

Profess i ona 1 : .3 

Other: None 

Project Description: 

Obj ect Ives : 

To reduce the turnaround tine seen by users of an 
off-line computer. 



Methods Enployed: 

Application of queueing theory to 
problem. Computer si^iulatlon studies 



the 



turna r^iund 



Major Findings: 

The turnaround time of an off-line computer running many 
short jobs can be drastically reduced by providing tv/o 
channels, one of v;hich is very high speed, and allowing 
the user to control the priority. 

A sophisticated ant I -overl oad technique has been 

developed to protect the short turnaround. The entire 

system is simple, inexpensive, and easy to install on 
existing hardware. 



159 



significance to Biomedical Research 

Ciomedical computing v/ork is often slov/ed down by the 
need for the user to v/a i t for turnaround'. The 
high-speed channel v;i 1 1 help tc alleviate this problen. 

Proposed Course: 

The algorithm vn 1 1 be coded for the I Br 360/370 and an 
experimental test obtained. 

Publ i cat i ons : 

Jones, T.L., "Load f^'anagenent for Short Turna rounri, " 
Proc. ACM GIGCOSIfi Second Annual Symposium, Jan. 1D73. 



160 



Serial Mo. DCPT 6.12 

1. Heuristics Laboratory 

2 . 

3. BethesHa 



PHS-NIH 

Individual Project Report 

July I, 1972 through June 30,. 1073 



Project Title: Radiation Treatment Plannin? 

Previous Serial Number: 6.7 

Principal Investigator: Louis Nodes, Ph.D. 



Other Investigators 
Cooperating Units: 
Man Years 



Frederick L. Faw., flCi 
Radiation Branch, NCI 



Total : 


.2 


Profess ional : 


.2 


Other: 


None 



Project Description: 

Object i ves : 

A proper course to brinf^ to ultimate widespread use of 
the system developed last year for interactive 
optimization of external bear^ radiation treatment 
pi ann i ng. 

Methods Employed: 

Consulting with various experts in this fiel-i. 
Submission for publication. Filinr; an Invention 
report with the NIH Patent Office. 

Major Findings: 

The sophisticated use of computers Is advancing slc\/ly 
but surely in radiation therapy. \.'e estimate there 
vjlll be a demand for a system like this in the near 
future. 



161 



Significance to Bionedical Research: 

Can lead to a najor inorovonent in radiatlrn theraoy, 
allowing therapists nore possibilities in plannin?: 
t reatment . 

Proposed Course: 

Publication is expected to stir^ulate sone denan^ for 
such a system. A governnent patent can be used to 
license a nanufacturer as an incentive to produce a 
marketable version. 

Publ i ca t i ons : 

"Semi -autonat ic Optinizaticn of External Bean 
Radiation Treatment Planning;" accepted by Radiolo'?y, 
subject to revision. Revised. 



162 



Serial No. DCPT 6.13 

1, Heuristics Laboratory 

2. 

3. Bethesda 



PHS-NIM 
individual Project Report 
July 1, 1972 through June 30^ 1973 



Project Title: Heuristic Methods for 

Chemical Structure Searchint^ 



Previous Serial Nunber: 
Principal Investigator 
Other Investigators: 

Cooperating Units: 
Man Years 



None 

Louis Modes, Ph.D. 

Alfred P. Feldman^ Walter Reed Army 
Institute of Research 

\7RAIR l,'alter Reed Army Medical Center 



Total : 


.5 


Profess ional : 


.5 


Other: 


None 



Project Description: 

Obj ect I ves : 

To develop a system for searching by substructure large 
files of at least a million structures In such a v/ay 
that it can be modified when efficiency deteriorates 
because of the addition of new sections to the file or 
when queries uncover defects in the screening process. 

Methods Employed: 

A preliminary system was designed using hash codes to 
speed up search so that an interactive system becomes 
poss i b 1 e . 

Major Findings: 

A hash coded substructure search system has been 
designed. Previously hash codes had been used mainly 
for f ul 1 St ructure searches. This search system is 
capable of finding more unusual structures faster and 
takes advantage of the heterogerene i ty of chemical 
st ructures. 

■ 163 



significance to D loriecl ical Research: 

Nev; compounds are being Introduced at the rate of about 
200,000 a year. It v/il) be necessary to face the 
proolem of substructure searching for large files« 

Proposed Course: 

The plan is to investigate the developnent of screens 
based on file statistics. This st^udy should take at 
least a year. 



164 



Serial No. DCRT 6.1U 

1. Heuristics Laboratory 

2. 

3. Bethesda 



PHS-NIH 
Individual Project Report 
July 1, 1972 through June 30, 1973 



R-oject Title : Computer Model of the Human Menstrual Cycle 

Previous Serial Numbers None 

Principle Investigators: John K. Dixon Ph.D. 

Charles Cargille, f'.D. 

Other Investigators: None 

Cooperating units: NICHD 

Man Years: .h 

Project Description: 

Object i ves : 

The development of a Dynamic Computer model of the 
hormone interactions v/hich are involved in the menstrual 
cycle of the human female. 

Methods Employed: 

The model is expressed as a set of coupled differential 
equations describing the hormone levels, grov/th and 
decay rates of organs, and transformations of various 
biological structures. This model is then compiled by 
the DYNAMO compiler and run from given starting 
conditions. The results of the run are then compared 
with experimental measurements v/hich have been published 
(such as curves of hormone concentration vs time). If 
serious discrepancies are noted then the model Is 
refined to obtain closer agreement. 

Major Findings: 

1, A model has been developed v/hlch appears to be more 
elaborate and more accurate than previous models. 



165 



2. A mechanism fcr the phenonenon of atresia of 
follicles has been proposed. The correct nechanisn is 
presently not knov;n. This proposed mechanism was 
discovered as a byproduct of modeling and appears to be 
a logical necessity from published experimental data. 

Significance to Biomedical Research: 

The construction of a good computer model forces one 
to bring together published information from a 
wide variety of sources and fit it into a logical 
and consistent structure. This procedure will 
(and has) reveal gaps in our knov/ledge and sus':;est 
critical experiments to fill in those gaps. This 
procedure may also suggest v/ays to fill in some 
gaps by logical inference or extrapolation from 
known facts. In addition this method may also 
reveal contradictions in published experimental 
results or in i nterpretat i nns of those results 
(this has occured in one minor case). 

When the model is sufficiently complete and vjel 1 
validated it vj\]] serve as a compact statement of 
existing knovJedge about the human reproductive hormone 
system. 

It may even be possible to discover new properties 
of the human system by experimenting with the model. 
Such a new property could then be verified by 
experimental study of humans or animals. 

Thus there are tvio major elements in tlie significance 
of this vjork. First, knowledge about the human female 
hormone system v.'ill be extended. Second, the value 
of dynamic modeling as a tool for understanding complex 
biological systems will be demonstrated. If this v/ork 
proves valuable further applications of systems dynamics 
to biological systems will be stimulated. 

Proposed Course: 

A paper on the present (preliminary) model is nov/ being 
written. The model will then be extended and made more 

accurate. Experiments on dynamic stability will be 
carried out. Then random elements v/i 1 1 be 
introduced and statistical data will be recorded. 
Later reports can be published in three places: 

1. The Endocrinology Journals 

2. The Computer-^''ed i ca 1 Journals 

3. The Computer Modeling Journals 



166 



■July 1, 19/2 through June 30, 19 73 

PHS-NIH 

Division of Computer Research and Technology 

Summary of Branch Activities DCRT 

Data Management Branch J. Emmett V/ard 

Branch Chief 

I . SUMMARY 

During F.Y. 1973, the Data Management Branch (DMB) continued 
its support of the MIH mission by adding significant extensions 
and programming support for several new NIH projects. In 
addition DMB continued its support of the Inquiry and Reporting 
System (IRS), the Mathematical and Statistical Program Library 
and various other types of software. 

Under the direction of Ms. Judith S. Prewitt, three principle 
program-projects have been undertaken during fiscal 1972-1973. 
Two of these are research programs concerned with medical pattern 
analysis^ characterization and representation, and decision- 
making. The third involves consultative and educational services 
for DCRT, other NIH institutes, other Federal agencies, and 
professional societies, and serving on advisory committees 
on Automated Cytology and Diagnostic Radiology for NCI. 

Problems of bio-medical Importance and Interest have tended 
to become more and more dependent on the analysis of large and 
variegated data bases. The more medical data that is gathered, 
the more necessary it becomes to Infer and understand patterns 
In the data, to categorize the data in classes, and to model the 
underlying processes v/hlch are generating the data. Application 
of pattern recognition methodologies has produced relatively 
few breakthroughs in real and pressing biomedical problems. 
Despite some notable contributions in picture processing 
machine diagnosis, and waveform analysis, there is no coherent 
theory of pattern analysis and recognition, merely a collage of 
heuristic, statistical, and linguistic approaches. The desir- 
ability of comparative evaluation of methodologies and assess- 
ment of their relevance to practical biomedical problems, in 
parallel with new theoretical studies, is poignant. Tovyard this 
end, the DMB is developing mathematical and computational tools 
which encourage and enable facile, creative interaction vnth 
real data, and feedback between modeling and data analysis. 
These facilities stress exploratory pattern analysis, 
characterization, classification, and modeling, v/ith 
supportive interactive graphics. They strive toward a 



167 



decree of flexibility and expandability sufficient to 
meet the denands of p;enerality or broad applicability 
in the experimental domain. 

Tv70 principle research-oriented projects, plus several 
minor projects, have served as specific vehicles for 
carryin?^ out these objectives. The principle research 
profirams have involved the conception and inp 1 enen tat ion 
of innovative nathemat ical and information science 
techniques, as v/ell as the application of appropriate 
existing contemporary methods. One of the programs 
focusses on medical diagnosis; the other is concerned 
v/i th prognosis-optimizing therapy. A collaborative project 
on quantitative characterization of bladder epithelium by 
digital computer has been Initiated at the request of the 
National Bladder Cancer Task Force of '!C I . The goal of 
the project Is to lay a quantitative foundation for biopsy- 
patiiology, v;hich Is the reference point for decisions 
on exfoliative cytologlcal material, obtained non- 
invaslvely., The data base consists of actual stained 
biopsy specimens which are imaged by and digitized by 
microscanner (JPL facilities). Specimens cover the range 
from normal epithelium to anaplastic invasive carcinoma. 
The quantitative characterization formalizes morphological 
notions of progressive disorganization Internal to 
individual cells, and external, between cells In context. 
An Integrated system of expandable interactive programs 
for the PDP-10 Is being developed to support this, and 
related diagnostic projects involving pictorial data 
files and biomedical pattern recognition. (1) The PFfiP 
system accomodates data management requirements for image 
processing, and provides means for mathematical and 
logical operations on scalars, matrices, pictures, and 
strings, using functions, relations and operators inter- 
actively. These facilities lead to picture segmentation 
and feature extraction and tissue characterization. (2) 
The SLAB system accommodates pattern Information pro- 
cessing and decision-making; it provides means for 
Interactively defining sub-sets of objects and relations 
among them In terms of attributes of members, and for 
making discriminations. Both programming systems are 
deliberately being designed to be of general utility In 
biomedical pattern processing. 

During the coming fiscal year, PEEP and SLAB vn 1 1 
acquire additional capabilities for pattern recognition. 
Specifically, (1) picture segmentation programs and 
vectorial and graph-theoretic representations for 
tissues (2) graphics to assist in pattern recognition 
and mul t I-vari ate statistics, and (5) classification 
and clustering programs will be developed. 



168 



The second research project program concerns 
prognosis-optimizing patient management/ using radiation 
treatment planning as a model system. Optimization is 
based on the use of mathematical models of response to 
therapy, and an objective function which expresses the 
expected therapeutic benefit as a function of treatment 
variables and tissue response parameters. Specifically, 
cell kinetics models for the differential radio-response 
of normal tissue and cancer are postulated, and radio- 
biological 1 y-motlvated mathematical definitions for cure, 
control, and normal tissue damage are given'. Using non- 
linear programming, the configuration of external 
radiation sources which (1) maximizes the expected 
therapeutic benefit (e.g. the probability of uncom- 
plicated cure), (2) ensures uniform post-treatment 
local control of the tumor, and (3) minimum total radia- 
tion exposure. Is determined. 

During the coming fiscal year, the capability of the 
program to handle external beam Information will be 
extended. Sensitivity experiments to determine model 
robustness, and the incremental gain of radio-biologi- 
cally optimized therapy over contemporary practices will 
be conducted. 

Mr. William D. Vincent Is directing two program 
projects In the Clinical Center. One project has the 
objective of integrating the current five departmental data 
bases (Clinical Laboratory, Blood Bank, Admission, Discharge 
Diagnosis and Anatomic Pathology). The intent here is to 
provide investigators with an easy to use retrieval 
package for interrogating this data. The files will be 
made available for all Clinical Center Inpatients and 
selected outpatients. This means that a method must be 
devised for adding new patients to the file and purging 
inactive patients from it. 

Data elements to be made available for update and 
retrieval have been designated. Development of a method 
for linking the files using the facilities available in 
IBM's Customer Information Control System (CiCS) is 
currently being defined. 

A second project involves the development of a data 
collection and retrieval system for the Meningeal 
Leukemia Study, in collaboration with Dr. Brigid G. 
Leventhal, NCI. Currently Mr. Vincent is defining the data 
elements needed for the study, determining the most feasible 
method of collecting and digitizing the data elements, and 
designing and developing a data collection and Retrieval 
System. 



169 



The following is a summary of those systems which 
were developed by the various sections in the branch and^ 
where appropriate, an abstract of those projects which 
demand a more detailed explanation. These projects can 
be summarized best by organization structure as follows. 

Documentation and Systems Support Section 

1. Clinical Center Data Processing Support 

Programs have been written to convert the active 
laboratory data files (BETA) from the CDC 3200 and 
establish procedures and routines for inquiry and pro- 
cessing of these data. Data relating to laboratory tests are 
updated weekly to provide current files to CC Investi- 
gators. Some general retrieval capabilities have been 
added to the process for the convenience of Clinical 
Center Investigators. These include retrievals by 
patient number and retrievals using test result 
parameters. The current process is quite expensive and 
DMB, in collaboration with the Office of Clinical and 
Management System (OCAMS), is looking Into methodologies 
for reducing these costs. 

2. Case 1972 Reports 

This preparation of approximately 96 different listings, 
tabulations and ranking tables Is developed annually for 
eventual publication by the OD/ADPPE/RA. The DMB Is 
currently developing generative programs which will eliminate 
the need for continuous modifications of existing programs 
to meet new needs. This approach will be production tested 
this year. 



3. 



Trends in Graduate Enrollment and Ph.D. Output In 
Selected Science and Health Professional Fields. ( 
1962 thru 1969-1970. ) 

Redesign of Arms Personnel System 



1961- 



The major thrust of this system will be directed at 
Improving the data collection and retrieval process by 
providing the 22 NIH personnel points with a terminal 
based input and retrieval capability. In addition, this 
system will attempt to interface in a digitized manner 
with the expected new DHEW system. 

The system should be fully operational by January 
197U. 



170 



5. Support for Committee Management Office, OD. 

6. VItam!n C Study. 

7. Clinical Center Census Reports. 

Using the Admissions and Discharge System Data Base, 
the DMB provided the Clinical Center with a complete set 
of daily, weekly, monthly^ etc. census reports. This 
resulted in a significant savings in man-time along with 
earlier reporting of this Information. 

8. Physician Authority List System, 

This system was developed to produce moc!c-ups of NIH 
Form 101 (record of participation In patiarrf care) for annual 
verification and to produce a directory of all physicians on 
patient care and mailing labels for ail physicians In the 
directory. 

In addition the system maintains an alumni (ae) file, 
produces a directory of all alumni (ae) and produces mailing 
labels for all alumni(ae). 

9. Collection of Data on Russian Scientists, Doctors and 

Medical Institutions for Fogarty International Center 

10. MHL! Information System Phase I! 

The intent of this project is to develop an MIS for 
operational planning that can be linked with the NHL! 
Grants and Contracts System. 

11. Program Monitoring System (PMS) 

In collaboration with NCI, the DMB provided a data 
retrieval system for all contracts In the Carcinogenesis, 
Viral Oncology and Demography areas. 

12. Pima Indian Study, PHS, Phoenix, Arizona. 

This project provides basic computer support for th® 
Pima Indian Diabetes Study in Phoenix, 



171 



13. Case 1972 Data Preparation - NSF 

lit. Opportunity Skills System - ODA, P 

15. Patient Registry Book and Plate Making - CC 

15. Diagnostic Radiology Department Case File - CC 

17. Fogarty International Center - FIC 

18. NICHD Grants System - iilCHD 

19. Computer-assisted Electron Microscopy - NHLI 

20. Contractor Data System - NCI 

21. Lupers Data - NIAMD 



172 



Applied Systems Programming Section 

1. Type II Intervention Study. 

The system supplies all of the data storage, valida- 
tion, purification, monitoring, and reporting capabilities 
required In support of the Type II Intervention Study 
being conducted by the Lipid fletabolism Cranch, i<IILI. 
The reporting capabilities v/ere added this year and the 
DMB is malntining the present system. Analysis of the 
data should be defined some time during FY 71+. 

This system makes it possible for Lipid Metabolism 
Branch personnel to store and to subsequently do reporting 
and analyses on their patients with blood lipid disorders 
and to relate these analyses to other family members. 
It also provides them with listings used as an aid to 
the doctors for patient handling in the clinic. Support 
for the system and training of fJHLI personnel in its use 
is supplied by the DMB. 

2. Carcinogenesis Bioassay Data System 

This computerized data processing system makes it 
possible for the Office of the Associate Scientific 
Director for Carcinogenesis, NCI to control the 
data acquisition, input, purification, reporting and 
analysis of animal experiment data related to the study 
and Identification of various agents in order to deter- 
mine their carcinogenic affect or capacity. The Study 
includes many thousands of animals being tested under 
various contracts awarded by the IICI. Examples of output 
are a summary report, pathology report, survival curve graph, 
and weight curve graph. Output can be produced in hard 
copy or on Micro-fiche, for all, selected contracts, or 
even selected experiments of the study. The contractors 
can, upon request, be provided with a complete data base 
subset of their data. In addition, the 'branch supplies 
assistance to contractors before and after their studies 
are included In the system. 



173 



3. Autopsy Data -ysten. 

This systen makes it possible for Dr. Young of the 
Comparative Pathology Section, Division of Cancer Treat- 
ment/ .'CI to store, maintain and to subsequently do logical 
retrieval from a data base consisting of the extremely 
variable '"linical and Pathological Hiagnoslr. data in 
autopsies of cancer patients. The primary purpose of this 
study is to determi ne v/ha t, if any, relationships exist 
bet\/een the use of specific agents and the occurrence of 
designated lesions. The computer assistance provider! by 
the i'ilP includes CPS data' acquisition, automatic 3';0P 
encoding of clinical and pathological diai^nosls, error 
correction, edit/update, a fixed format report, and a 
completely generated COf^OL retrieval program executed 
via T30, v/hich provides almost complete flexibility in 
composing quick queries. 

h. Clinical Center .;iood ^ank Computerized [.ata ''anap-enent 
3 y s t em . 

The system is being designed to place all pertinent 
plood i'.ank data into a computer processable media making 
it more accessible by authorized investigators throughout 
the Clinical Center. This approach will Increase the value 
of the data, both from a statistical and reporting stand- 
point and v/i 1 1 assist in improving patient care and 
research. The systen v/ill Interact v/ith v/orking CCRE 
personnel, making their v/ork easier and less error prone, 
it v/ill monitor each blood product by unit from the time 
it is logged into the system until it is completely 
d 1 spensed . 

5. '!IH Appropriations and Obligations Information Cystem. 

This computerized data processing system provides the 
Office of Legislative Analysis, OD, MIH, v/ith the 
facilities to maintain a data base of f.'IH Appropriations 
and Obligations by total ."ilH, total 'WW minus Programs 
Transferred out, by activity within r.'IH and by activity 
within program. 

6. ;.'IH Personnel Accident Reporting System for the Office 
of Administrative Services, Protection and Safety 
'management. 

7. r.'IH Space flanagement System for Office of Administrative 
Services, Ceneral Services flanagement. 



174 



8. Medical Records Discharge Olagnosls System. 

The system provides the Clinical Center Medical 
Records Department with the ability to maintain a 
current, accurate data file of all Clinical Center 
patients' discharge diagnoses. It also supplies various 
useful reports and listings. 

9. NIH Internationa] Activities and Personnel Monitoring 
System for the Fogarty International Center, 

10. Type V intervention Study. 

This system provides Dr. Greenberg of the Lipid 
Metabolism Branch, NHLi with the capability of storing 
and purifying for analysis data collected by NHLI on 
patients with TYPE V lipid disorders and their first 

degree relatives. 

11. NHLI Personnel Data Management System 

The system Is designed to help the NHLI personnel 
office maintain an accurate, up-to-date data base con- 
sisting of all NHLI personnel. It provides a very flexible, 
easy-to-use query capability In addition to fixed, re- 
curring report processing modules. It Is hoped that 
this system can be integrated with or phased out by the 
central ARMS personm®! system in January, 197Jt. 

12. Mutant Data Update System for the Laboratory of Molecular 

Biology^ NINDS. 

13. Training Information System for the Personnel Office, 
Office of the Director, National Institute of Dental 
Research 

We are acting as consultant to NIDR personnel In 
their effort to Implement a modified version of DCRT's 
Training Information System. 

li». Population Research Scientist Registry. 

This system provides a method for the Center of Popula- 
tion Research, NICHD to record data on Information requests 
and applications. Programs have also been developed for 

retrieval purposes. 



175 



15. HL-A Data Maintenance System. 

The nature of Dr. Ro^^ent i ne ' s study for the !nmunolo?;y 
Branch of "ICI is to try to determine If there are any 
connections betvyeen specific :!L-A types and disease, and 
to Identify v/hat they are. 

16. !)CRT Personnel and Training; Systen. 

17. Leukocyte Infusion Update and Reporting Systen 
Leukemia Service^ Fledicine Cranch, :.C ! .- 

The system collects data and reports on Leukocyte 
Infusion information. The information collected includes 
specific infusion information, laboratory fol1ov/-up 
information, and infusion series summary Information. 
Data is extracted from tv/o related systems, the Cell 
Collection and the Donor Recipient Information Systen. 
These data are then combined v/ith the Leukocyte Infor- 
mation for other reports. 

18. Lipid Protein Study - Lipid i'.etabolism Branch, ;!!;L! 

19. ;.'HLI Personnel Data Management System ~ OD ;!HLI 

20. Personel Information System - Dept. of -^avy 

Plans for FY '74 Include continued support of the 
above mentioned projects with an emphasis being placed on 
balancing the work load to the available manpov/er while 
retaining a maximum degree of flexibility. To accomplish 
this^ The Applied Systems Programming Section v;i 1 1 try to 
continue to give heavy support to the developmental 
effort of three major projects and concurrently support 
the development of ten to fifteen less man-t ine-consumi ng 
projects. This v/i 1 1 also make It possible for us to 
continue to further our employees' career development 
by enrolling then in various training programs. 



176 



Scientific Application Section. 

1. Psychophysiological Measures 

The intent of this system was to develop a general- 
ized system for the Laboratory of Socio and Environmental 
Studies, NIMH, for the editing, storage, retrieval and 
statistical analysis of psychological and physiological 
measures. As a result of developmental research, 
different modules in the system underwent extensive re- 
vision. Of particular significance was the' area of 
Statistical Analysis. Subsequent efforts are anticipated 
to be in the area of integrating the computer processing 
with the laboratory processing of data. In addition, data 
updating capabilities are being considered. 

2. Etiology Smoking Test 

For the past three years Hazel ton Laboratory has been 
applying tobacco condensates to 13,500 mice. This pro- 
ject IS sponsored by the Etiology Branch of NCI. DMB has 
supplied the computer support to collect the data con- 
taining dosage, deaths, tumor observations, and 
classification. 

DMB has established a data base and provided an up- 
dating capability. In addition programs have been 
written to compute dally survival chl-squares and to pro- 
duce a number of Calcomp plots comparing test condensates 
with the standard and observed data against model - 
predicted data. Systems documentation was turned over to 
NCI for future processing. A second large group of mice 
is currently under examination. This second Phase data will 
be processed utilizing Phase 1 programs. 

3. Low Weight for Age Study. 

The objective of this study Is to define that group of 
women prone to delivering a baby with low birth weight for age, 
Analysis was done on a file of almost two million records 
which represents one-half of the births in the U.S. 
during 1968. This system supports the work of the 
Epidemiology Branch, NCI. 



177 



k. Data Base for Surgery Branch 

The Surgery Branch of the NCI had a need for a data 
collection and reporting system. This included the 
editing, storing and retrieving of all data related to re- 
search projects of the Surgery Branch, NCI. To date, the 
DMB has provided CPS prompt programs as well as an update, 
retrieval and reporting system for seven NCI protocols. 
Data files have been developed to include one file of 
common data and an associated auxiliary file. This rep- 
resents a file of all Surgery Branch Operations including 
those which have not been expanded into separate pro- 
tocols. Future considerations are the possible inte- 
f ration of two more protocols. Analysis of the data may 
ead to additional protocols for operations not under 
special investigation at present. _ 

5. Cancer Survival System 
NCI 

The Survival System was originally developed to 
support the End Results in Cancer Studies. Maintenance 
and Improvement of the system Is the primary goal. 

During the year several requests were made for copies 
of the system. In addition there has been Increased use 
of the system by elements outside the Cancer End Results 
Group. The Survival System and a set of post-processing 
programs were added to the Survival Library for use by the 
NIH community. The new post processing programs included 
a group of Chi-square programs for comparing pairs of 
survival tables, a program to produce Calcomp plots of 
survival tables, and a program to produce summaries of 
the tables. 

6. Current Awareness Search of Clinical Biological Activities 
(CBAC) 

The biweekly Issue of the CBAC data Is received from 
the Chemical Abstracts Service (CAS) in Columbus, Ohio. 
These issues are used to supply the NIH community with 
information on current articles related to the chemical- 
biological area. The CAS format for distribution was 
changed In January; this precipitated a major rewrite of 
the reformat program, which precedes the search of the 
CBAC data base. Significant savings have been realized 
by coding the new reformat program in assembly language. 

Future considerations include the addition of new data 
bases of NIH interest. 



178 



7. Aortic Valvular Disease 

In collaboration with Dr. Walter Henry, NHLI, the DMB 
is providing programming support for a study which will 
attempt to define prospective pre-operat i ve risk for heart 
valve replacement patients. The DMB has provided data 
collection and file maintenance programs to support this 
effort and hopes in the next fiscal year to provide a set 
of interactive analysis programs. 

8. Cutting Oil Study 

This system supports the efforts of the National 
Institute of Occupational Safety and l^ealth, HSMHA. It 
is a study relating job type to mortality and morbidity 
of cutting oil workers. A data base has been established, 
and a survival analysis program similiar to that developed 
for the Steelworkers Project will be v/ritten. 

9. PDP-10 Program Conversion 

Several benchmark projects v^/ere converted from outside 
comnercial time sharing systems to run on the DCRT-PDP-IO 
computer. Programs and documentation v;ere released for 
technical evaluation. 

10. Dermatogl yphlcs Data ~ fJlCHD 

11. ,"iul 1 1 phasic Zone Electrophoresis Programs - NICHD 

12. fiyasthenia Gravis - NEI 

13. Substructure Search - NCI 

Ik. Kidney Dialysis Contract - HI AMD 

15. Occupational Studies - NIOSH 

16. Enforcement and Heneral Counsel - EPA 

17. ALC/SDF Reformat Program - Chemical Abstract System 



179 



Math/Gtat Section 

1. Productivity measures at :iiH. 

This Is a completed project for the 0/DA, :;IH in- 
volving the evaluation of man-year, output and input v.'ith 
deflator indices to compute, based on 1957 measured out- 
put, productivity measures of activities in terms of man 
years, v/ases, unit labor cost and unit labor requirement. 

2. Analysis of Data of Inbrer! .''ice and lutritlonal Treatments 

In support of jRS programs v/ere required to evaluate 
data involving rations containjn,^: two levels of crude 
protein and three levels of fat which were fed to four 
strains of mice. Split plot design and least square 
analysis programs have been run to evaluate the effects 
upon formula rations on reproductive performance and 
reproductive performance with different physical charac- 
teristics. 'Jev; programming Is underway to evaluate re- 
productive performance with respect to weight of all 
v/eanling mice. This study v/i 1 1 also expand to analyze 
reproductive performance of the 'ilHAC foxhound colony. 

3. Interview Scheduling. 

The Clinical Center requested major modifications to 
the original specifications of the Clinical Associate 
Interviev/ Scheduling System. Systems modifications to 
make programs compatible with the new F0RTRA[!-^1 compiler 
and I/O packages have also been completed. This modified 
system was successfully executed, producing an optimized 
schedule of interview periods for both the applicants 
(for associate program) and the I n terv I ev/er s (scientific 
staff at NItl). 



180 



k. Evaluation of CC Heart Data. 

The Initial objective of providing a massive storage 
and retrieval capability for chemistry data of the 
Endocrinology Branch, NHLI has been accomplished. Statis- 
tical programs have been prepared to evaluate time trends, 
to produce chemical balances, to calculate degree of 
saturation of urine with respect to brushite, and to study 
formation of renal stone of calcium phosphate origin. 
Statistical evaluation is under way on chemical data com- 
bined with patient demographic Information -and fourler 
analysis on blood pressure data. The final objective of 
this study will be to develop a capability to produce 
statistical output and evaluations that would aid 
scientists to study the data with effective scientific 
tools. 

5. Retrospective Study of Aortic Valve replacement. 

A set of programs have been completed to edit data on 
700 patients who underwent open heart operations. Simple 
statistical parameters have also been computed to help 
study statistical approaches that would evaluate post- 
operative results of aortic valve replacement and define 
the morbidity and mortality of patients undergoing open 
heart operations. 

6. Lipid Protein Study. 

In support of NHLI, this study involves evaluation of 
data on Type II kindreds whose propositi were the first 
120 Individuals found to have Type il hyperlipoproteinemia 
at the Clinical Center. To date data description statis- 
tics have been computed and analysis of variance runs have 
been completed. Under way are multivariate variance 
analyses runs to evaluate effects such as sex and age 
categories In addition to the group differences. Evalua- 
tions are sought for the significance of Type IIA vis-a- 
vis Type I IB. 

7. Cross Transfusion Studies. 

This project Involves analysis of basic transfusion 
experiment data collected from an ongoing study in the 
NHLI. The first phase of this study involved developing 
and Implementing algorithms that would calculate renal 
functions. The second phase will include the statistical 
evaluations of renal functions once the calculations are 
proven satisfactory. 



181 



8. Evaluation of Psychiatric Ward Value Questionnaire. 

This ongoing project Involves evaluation of data to 
reveal dimensions which appear of practical and theo- 
retical interest for characterizing ward value orien- 
tation. There is also considerable interest by NIMH in 
assessing the statistical reliability and validity of the 
factor analytic dimensions. Test scoring programs which 
edit and transform the raw response items and compute a 
subject's scale have been implemented. A new phase of 
this study is beginning to emerge and this project could 
become very large. Discrlmlmant analysis and factor 
analysis programs are the principle analytical tools for 
the present evaluations. 

9. Identity Cognition. ^ 

NIMH is conducting a longitudinal study of cognitive and 
self Image development among individuals In several con- 
trasting adolescent samples. Even though the principle 
Investigator for this study left NIH, major modifications 
have been made to the programs for basic analyses of self- 
image intercorrelat Ions at single points and, between 
testing, Intercorrel atlons of segments over time and tests 
for their significance. Variance analyses are also being 
provided for the results of the cognitive tests. Investi- 
gators will/ hopefully, derive empirical statements which 
will lead to more precise measurement of Identity develop- 
ments and associated cognitive processes In diverse and 
contrasting populations. 

10. Analysis of Clinical Research Data. 

This NIMH study involves the evaluation of patient 
behavioral data and patient data. To date, all analysis 
has involved behavioral data including drug and shock 
data. Programming tasks have involved editing and trans- 
forming behavioral data. Factor and time trend analysis 
is being used to detect behavioral changes. 



182 



11. Problt and Loglt. 

This DBS project involves evaluation of different com- 
putational methods In an effort to refine the assessment 
of bfologScs products. The progranmlng task Is completed; 
it computes problts and logits with options to compute 
individual or pooled slopes/ with or without conversion 
of input data. Options In the programs Include tests for % 
parallelism and relative potency. Further modifications 
and generalizations are expected as new developments occur. 

12. Evaluation of Pertussis Potency Tests. 
DBS 

This project Involves the evaluation of Pertussis 
vaccine potency as performed at DBS. Programs performing 
problt analysis have been extensively utilized in this 
study. However, additional programming Is underway to 
determine If there are seasonal variations In the test as 
evidenced by cyclic changes in standard vaccines or changes 
in the Immunity of animals receiving standard vaccines. 



183 



Software Support Section 

1. RMAG MACRO Generator and Program Generation 

RMAG (Recursive Macro Actuated Generator) is a pro- 
gramming language which can generate source coding for 
any language for which it has the character set. It is 
independent of running computer, target computer, and 
target language. The uses of RMAG range from that of a 
synthesizing editor used to produce di f f i cu 1 t-to-wr i te 
fragments of source language code to that of a computer- 
system generator. 

RMAG is the main tool used in writing the source- 
language generators produced by DMB. These source 
language generators have reached a level of sophistication 
in which complete single file systems for data file 
creation, maintenance, validation and retrieval are being 
generated in package form, ready for immediate use. This 
common approach to handling typical research data manage- 
ment problems has resulted in significant reductions in 
programming costs during the past year. This success of 
program generation Sn the NIH environment has come to the 
attention of several other government agencies in the 
Washington area and the DMB is in the process now of 
making its software available to them. 

RMAG has also been modified to run interactively under 
IBM's Time Sharing Option and a new capability, caHed 
Checker, has been added to interactively writing and 
testing RMAG routines. 

2. Medical Monitoring Program 

This system was designed to support the Emergency 
Virus Isolation Facility, NCI. Employee physical, accident 
and virus exposure data are collected regularly and periodic 
reports are produced for perusal by management personnel. 
The current data collection process Is being modified to 
enable interactive Input by NCI personnel on site. This 
will provide much better data security for personnel, who 
work In the EVIF. 



184 



PUBLIC HEALTH SERVICE - NATIONAL INSTITUTES OF HEALTH 

DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 

Report of Program Activities 

July 1, 1973 through June 30, 1974. 



ANNUAL REPORT 
TABLE OF CONTENTS 

Page 

Director's Summary ------------------__,_ ± 

Summary of the Associate Director ------------ 9 

Summary of the Assistant Director ------------ n 

Computer Center Branch ------------------- 13 

Laboratory of Applied Studies -----------___- 2I 

Computer Systems Laboratory ---------------- 27 

Physical Sciences Laboratory ----------------- 41 

Heuristics Laboratory ------------------- 47 

Data Management Branch ------------------- 53 



DCRT Annual Report 
July 1, 1973 through June 30, 1974 
Office of the Director 

It has been ten years since DHEW approved the concept of a new NIH division to 
cope with computers and to help reap the many benefits anticipated from "the 
broad application of this new technology to research in the life sciences." 
The FY1974 activities of the Division of Computer Research and Technology, set 
forth in detail by the laboratories and branches in this annual report, show 
the concept was prophetic and is still essentially sound. 

This Director's Summary highlights these activities briefly and speaks to the 
future, wherein lie the opportunities and problems for computing at NIH. Four 
figures provide a visual FY74 synopsis of the size and functions of the DCRT 
components and the management and revolving fund expenditures, as well as a 
multi-year overview of growth in the NIH computer centers task load and in the 
revolving fund services rendered. 

The Past Year 

Computer Center Branch continues to provide NIH with responsive, reliable com- 
puting services and facilities of unexcelled quality and diversity. Use of 
both central systems (the IBM 360/370 system and the Digital Equipment Corpor- 
ation (DEC) PDP-10 system) grew (now more than 200,000 jobs per month). Costs 
per user job were cut 30% or more. Major hardware improvements were installed 
on the DEC system and planned for the IBM system. Major software (systems pro- 
grams) improvements were made on both. 

What is important to emphasize here is the fact that growth in quality, quan- 
tity, sophistication and efficiency of the system and in economy and ease for 
the users has continued year after year. The ability to maintain and improve 
integrated man-machine-software systems of this magnitude is significant and 
of fundamental importance to NIH program activities. A dramatic example of the 
breadth of system capability is its application to store and analyze x-ray 
crystal lographic data then to depict macromolecular configurations in 3D stereo, 
and finally to create by computer output an inexpensive microfilm stereo atlas 
of protein molecules for biochemists throughout the world. (This can be exam- 
ined with a simply polarized viewer on conventional microfiche readers..) 

The Data Management Branch is equally ■'mpressive in its area, i.e., designing 
and creating couip-jter programs to rrianage and analyze data from scientific and 
administrative activities throughout NIH. In thii way the Branch transforms 
the basic computing system power into useful work. The "application programs" 
are then owned and operated by the NIH scientific or admiriistrative staff wno 
heve responsibility for the data. (The Branch also develops "software support'" 
tools to speed the construction of their applications programs.^ 

Even though the DMB report is impressive, listing almost a hundred projects in 
FY74 for the great majority of NIH institutes and divisions, it understates the 
Branch's impact on NIH activities. It does not depict the many applications 
built in previous years and still functioning well throughout NIH. The cumula- 
tive effect over the last several years has greatly strengthened the useful 
application of computers in virtually all parts of NIH. 



The Computer Systems Laboratory must be viewed from a different perspective. 
The laboratory develops computer systems for those activities at NIH which need { 
something beyond the power of the NIH central facility. As the CSL report notes, 
these are mostly applications wherein the requirements for data acquisition, 
analysis and response require a dedicated or locary shared computer. 

The CSL domain therefore includes the design, procurement, development, testing 
and installation of smaller computers and of their electronic interfaces, sys- 
tems software and applications programs. The results are not measured in hun- 
dreds or hundreds of thousands of jobs per year but rather a few projects which i 
each typically span several years from genesis to completion. ' 

The CSL also has its cumulative impact through systems completed in prior years. 
Much of the CSL effort in FY74, as in the previous year, went into projects in 
support of clinical care, particularly the intensive care unit for heart surgery 
patients and the NIH clinical laboratories. Two new projects begun this year 
demonstrate that there continue to be entirely new applications for electronic 
engineering and computer hardware expertise in NIH biomedical research. 

The Laboratory of Applifjd Studies represents another part of the spectrum of 
DCRT disciplines, which extends across computer technology, engineering, applied 
mathematics and their uses in biomedicine at NIH. The LAS report notes that its 
primary emphasis is on direct application of mathematics and statistics. In 
addition, LAS does research in these fields. Its bibliography shows that both 
endeavors are productive. 

Indeed, the LAS work illustrates within the DCRT the breadth of computing 
throughout the NIH. In FY74 the LAS staff worked in electrocardiology, nuclear 
medicine, biochemistry, neural modeling (computer and mathematical) microcircu- 
latory modeling, hardware and software systems for data analysis and a variety 
of other projects. These involved extensive collaboration with other parts of 
DCRT as well as with many NIH laboratories and branches and with coworkers out- 
side of the NIH. They use computing resources selectively but in a fashion 
which demonstrates need for the full breadth of disciplines originally speci- 
fied for the division a decade ago. 

The Phy s ical Sciences Laboratory also applies mathematics broadly to the exami- 
nation and elijcida*:ion of biomedical phenomena at a fundamental level. Its work 
on the theory of intermolecular forces in biochemical aggregation moved forward 
accompctnied by the development of new experimental techniques to confirm the 
theoretical insights. Advances in nuclear magnetic resonance instr-umentation, 
with computer cont.'ol and analysis, led to new theoretical work on polypeptide 
coiling and new discoveries about the interaction of subatomic particles under 
varying biochemical configuration. 

The PSL expertise carried to many other areas such as the dynamics of macro- 
molecules, microorganisms and blood elements, the interpretation of kinetic 
biochemical analysis and the design of clinical trials. The PSL bibliography 
speaks to the productive collaboration of this group with investigators in 
other NIH laboratories. 



The Heuri5tics Laboratory applied computing methodology to information handling 
and analysis for a wide variety of topics, extending its fundamental work on 
theorem proving to computer program verification and its pattern recognition 
efforts to liver disease cases, mass spectral data and tissue- typing screens. 
Its previous collaborative efforts in developir:) mass spectral data file 
searching systems now have international recognition and sponsorship and it 
began collaboration on design of a more flexible and powerful system for the 
Walter Reed (Army) chemical substructure search system. 

The HL work on data analysis and modeling saw many new users for the MLAB system, 
a more sophisticated approach to linear multiple regression analysis, and the 
development of an improved simulation model for an NICHD menstrual cycle model. 
Its bibliography also testifies to active collaboration with investigators in 
other NIH laboratories. 

The Present 

A prime influence on the DCRT at present are the personnel limitations imposed 
as part of the ceilings placed upon the NIH. Cuts in budgeted and actual DCRT 
positions have come progressively, during the yery years when its new Labora- 
tories and Branches matured, became more productive and engaged fully with the 
rest of the NIH. The division absorbed these cuts and the remaining staff over- 
came them with increased individual effort and more effective use of computa- 
tional power and sophistication. But in FY74 perennial retrenchment finally 
forced a reorganization and withdrawal from a DCRT commitment to those funda- 
mental efforts embodied in its Heuristics Laboratory. 

The particular irony of such a necessity can only be appreciated by viewing 
the course forseen by those who created the division a decade ago. The 
Steering Committee in 1963 declared: "The measure of the Division's success 
will be the extent to which it stimulates the introduction of ADP methodology 
into on-going programs." Computing facilities and services, data processing 
capabilities, computer system engineering and the application of modern mathe- 
matics to biomedicine have become a functional reality in the daily life of 
the NIH. Now is the very time when the mathematical and information hcindling 
research base should start making its greatest contribution to the uiH mission, 
but U is the very time at which this base had to be cut. 

The Future 

The DCRT commitment is fundamental support of effective computing in NIH pro- 
grams. To this end and under the reality of .-774 cuts, a new DCRT laboratory 
was initiated this year, drawing expertise from several other parts of the 
division. This laboratory will contain a multidisciplinary nucleus of pro- 
fessionals in contact with the developments outside NIH, and like other DCRT 
laboratories it will do research. But its prime responsibility will be to 
bring modern computer based capabilities in data analysis, mathematical statis- 
tics and other mathematical methods closer to biomedicine in the NIH programs, 
both by its own activities and by support of biometry in other parts of NIH. 

At this time, moreover, extensive involvement of the DCRT program as a whole 
in NIH program activities forbids any further reduction of effort in any DCRT 



laboratories and branches, if the fundamental DCRT commitment is to continue 
without loss of effectiveness and momentum. The original 1963 prospectus for 
"the computer division" envisioned growth from 335 employees in FY1965 to 680 
in FY1969. This prophecy may have been (.snrealistic, but it foretold the need 
for some new activity as the promise of computing became fulfilled in NIH pro- 
grams. Computing is now making steady strides in medicine, biomedical research 
and management. The fundamental strategy for the NIH must be to defend its 
carefully developed computing base against further retrenchment or retrogressive 
influences. 

The danger appears nowhere more clearly than the threats to the basic NIH com- 
puter capabilities posed by some administrative philosophies at other Federal 
levels. The technical expertise and managerial acumen which developed the 
central NIH hardware/software systems has been unexcelled, if indeed matched, 
elsewhere in the Federal establishment. In large measure this is true because 
the NIH leadership understood the complexity of computing in the NIH environment 
and delegated responsibility to the level where the requisite technical exper- 
tise and managerial experience could be established. Continued knowledgeable 
support by the NIH leadership will be needed to prevent erosion of the strength 
and effectiveness by external administrative requirements which lag behind the 
current technology and are virtually inapplicable to the complex, multiproces- 
sing, multiprogramming, multidisciplinary NIH environment. 

The Steering Committee Report in 1963 identified this basic problem in its 
discussion of computing in the governmental setting: "Government installations, 
particularly Federal scientific laboratories, seem to have encountered more 
serious problems than either universities or industrial centers. Administrative 
policies have been formulated with undue consideration to the use of computers 
for business or administrative applications." In spite of recognized competence 
and leadership, the DCRT and NIH carry a continual burden of educating others 
about the NiH program requirements. 

The need is for establishment of the historical perspectives and the prospective 
understanding which will allow NIH on-going programs to continue to benefit 
fully from the technology now that it has been introduced. The Steering Commit- 
tee words in 1963 are again worth noting: "The present Committee was impressed 
by the fact that the conclusions it reached after protracted analysis, and which 
are embodied in this report, constitute, in terms of broad outlook a.d major 
thrust, another step in a continuum first ar;iculated almost a decade ago." 

Recognition of the continuum and its progress^ion is d^ important as appreciation 
of the accomplishments of a single year. Electronic computing i? a little more 
than halfway through its second decade at NIH. It is currently a very consider- 
able success due to previous foresight and enterprise. The degree of success 
today does not reduce the need for foresight today and enterprise tomorrow. 






«Jt>.;;' 














i 1 



7^ 

o 



-n 



^ 



n: 03 — < 
m o 

O 
O CO 



I^ 






4S 



I 






CT5 



;:3 















^ 



i 






i^ 



r 



^- 












t^ 



^ 



^ 



^^ 



O 

o 

I — 

5t: 

















i-> 




Ci. CN 






L' J ; i ' 


i^ 




— ' S 
Ci. 1 — 


IL. 




<C ^'^ 







<^ 


i^c.% 




tL> zj 


I- 


^ ^ 


c~^ 






r: c\ 


J., 


oFf; 




c_> 


^ 




f^ 



t— 


«j\ 


2 




LD 




^l-^ 


o 


J- LO 


o: 


<C CD 


i- 


Ci «^ 


U. 


^ 


(X 


< 


^ 











gs-f 









Csi 






Ui 






*— 


o 




ua 


t- 




CD 


ts\ 




Cs: 


o 




LL> 






1— 


t- 




13 


U. 




D. 


a. 










o 


crs 




o 







o 



o 

:> 

UN 



m^/^E95mi9 



o o 



OS o 

c o 



O Q 



— M tM 

on o M 
O o O 



M »^ W m V*« UJ tk) 

Q- «» o i-Ni j» a- cti 
O O Q C o o 



-^ 4^ jb, jk ji vn 
O K> -^ o- w O 



o o o o o o 



i I 



I I { I M 



c-i 



o 




^ 






6^1 















Summary of the Associate Director 
July 1, 1973 through June 30, 1974 

The DCRT Scientific and Technical Information Office (STIO) became involved 
by error in the departmental attack on public relations puffery. Fortunately 
the basic duties of the STIO, set forth in 1970, were clear and dispelled 
the error quickly. The new emphasis on "inquiries" and "reports" fits the 
established STIO functions. 

The implication that "information" is henceforth a dirty word in DHEW may pose 
semantic problems for DCRT. Computers are, after all, information processing 
machines; information processing, with or without computers, is a large part 
of the NIH job; and the STIO has a functional responsibility for "communi- 
cation" of information about the processing of biomedical data by computer 
based techniques. This function amounts to more than answering inquiries and 
generating reports. 

The STIO edited and published two DCRT Technical Reports during the year: 
No. 9, A Structured Assembly Language Source Program Generator and No. 10, 
The Carcinogenesis Bioassay Data System . Both describe systems developed 
within our Data Management Branch. The office set up a standing Review 
Committee, composed of representatives of each Lab and Branch, to support 
standards of style and substance in the Technical Report series. 

The office continued to receive and answer inquiries from outside NIH. In 
the main, these ask about medical applications, careers in computing and 
specific computer software or scientific information centers. The office 
arranged formal visits for two national organizations: The American Medical 
Writers Association and The National Association of Medical Explorers. To 
answer questions of visitors to the Computer Center and to other parts of 
DCRT, the STIO designed an exhibit in conjunction with Medical Arts and 
Photography, DRS, and installed it in the lobby between Buildings 12 and 12A. 
It contains short "reports" about the functions of the Division and its 
Laboratories and Branches. A fuller set of brief reports on DCRT components 
and activities is planned, partly for this exhibit but equally important for 
rapid and effective briefings of the many technical and managerial visitors 
who come to DCkT from many parts of the world and the Federal -government.. 

More extensive specific reports for tht NIH staff i^ave been suggested. The 
STIO d'^stribut'^d a bulletin of recent DCRT libv-ary acquisitions and tech- 
nical reports to in'IH scientists and administrators at the Lab and B; anch 
level. The items were a sample de"'iberately selected x.o explore interests 
in a wide range of managerial and scientific computing topics. Most responses 
were requests for technical materials which are already loutinely announced. 
The bulletin was discontinued and more direct approaches will be used to 
direct readers to DCRT library acquisitions and to find out the topics of 
major interest for special reports. 

As in previous years, the office was responsible for many recurring adminis- 
trative reports, clearances and archival functions in support of Division 
programs. 



The DCRT Library strengthened its collection of documents in computer science a 
and related fields, adding about 200 new books and 100 technical reports. " 
Five new periodicals were added and four were cancelled or ceased publication. 
The library continues as a unique resource for NIH and for many scientific 
or technical workers in the Washington area. Almost half of those who borrowed 
books were not DCRT staff. 

Library operations were enhanced by a new automated circulation record system, 
adapted from that developed for the EPA library. It has been used for a year 
and supports our circulation policy very successfully by permitting borrowers i 
to keep books for an extended period and at the same time quickly and effi- 
ciently locating books when they are needed by other users. Borrowers peri- 
odically receive a listing of materials checked out to them, making it easy 
for them to keep track of and return their individual holdings. 

The library added the NLM MEDLINE to its services on a trial basis. The 
response was promising, with vary active use during the first two months of 
1974 by medical students in the "Computers in Clinical Medicine" elective. 
A terminal is planned ir, the library when a telephone line becomes available. 
This should stimulate more use of the system. The STIO will undertake an 
evaluation of its utility as an information service in our environment. 

For the coming year we are considering (1) a survey to eliminate unused 
journals, (2) weeding the monograph collection to discard outdated materials, 
(3) further development of an in-house training program for the library 
assistant and (4) additional automation of selected library functions. 

The Computers in Clinical Medicine elective was given for the third year as 
part of the NIH Clinical Elective program for medical students. For the 
first time it incorporated a two day seminar with invited experts, Drs. 
Howard Bleich, Jerome Cox, William Yamamoto and Homer Warner. This addition 
was quite successful, in large part because of the high quality of the 
students. Several students from this and previous years intend to return 
to NIH for Associate positions. The elective does, however, place a con- 
siderable load on DCRT and other NIH staff, and it wiT; undergo a detai''cd 
evaluation and possible redesign before it is offered again. 



10 



Summary of the Assistant Director 
July 1, 1973 through June 30, 1974 

The Office of the Assistant Director, DCRT, provides four basic capabilities: 

1. It serves NIH as the focus for NIH-wide coordination of ADP 
policy matters. 

2. It serves NIH as a central point of contact with PHS, -the 
Office of the Secretary, other DHEW agencies, GSA and 0MB, 
relative to NIH ADP policy questions. 

3. It supports the Director of DCRT by providing a point of 
reference aad coordination to insure that DCRT's own ADP 
activities are consistent with NIH, PHS, 0/S, GSA and 0MB 
policy directions, and 

4. It supports the Director of DCRT by providing advice on 
resource acquisition and allocation necessary for DCRT to 
support the above functions. 

The coordination role is most often exercised in connection with the fact 
that all proposals for contracts or procurement actions involving ADP 
equipment, services, programming or design are cleared through this office 
prior to being executed. This provides a continuous opportunity to alert 
program or contract officials to opportunities to reduce costs and/or avoid 
duplications and to avoid difficulties with higher echelons. 

During FY74 extensive assistance was, for example, provided to ODA/NIH to 
develop an RFP, evaluate proposals and negotiate a $304,310.00 contract to 
develop a single comprehensive Material Management System for NIH. Our 
support of NCI's effort to develop a primary Cancer Research Center at 
Fort Derrick in Frederick, Maryland, continued. Significant unnecessary 
costs were avoided and capabilities improved by arranging for the prime 
contractor at Fort Detrick to himself develop an ADP capability in lieu of 
subcontracting. Assistance was provided to the National Library of Medicine 
in beginning the long process of obtaining PHS, OS and GSA approval for a 
major ADP hardware upgrade. 

A major undertaking started during FY73 but continuing into FY74 V7as the 
technical and management leadership for NIH's effort to develop its second 
comprehensive ADP Plan. This plan, now an annual effort, attempts to lay 
out a two year projection for ADP equipment, manpower, contractr, and 
business data systems for all components of NIH. This planning process 
creates an orderly opportunity for ADP users to take stock of their goals 
and accomplishments. The second annual plan covered NIH ADP efforts 
exceeding $31 million for FY74 . An NIH wide study of systems haudling 
data coiicerning NIH grants was a direct result of apparent duplication and 
overlaps detected during the planning cycle. Work on the third annual plan, 
that for FY75, was underway at the close of the period covered by this 
report. 

11 



Page two 

In serving as a central point of contact for NIH on ADP related matters 
with PHS, DREW, GSA, 0MB, etc., a large number of NIH research and research 
support staff members are spared the agony of becoming expert in the many 
nuances of ADP management regulations. Since these regulations are generally 
written from a second generation business data processing point of view, a 
thorough understanding of their purpose and operation often allows beneficial 
interpretations of their application to the NIH research environment. 

During FY74 the PHS was recreated including an ADP Policy Office. The 
Assistant Director continues to provide extensive assistance to PHS to 
attempt to mold that office into an entity which can be helpful to NIH 
rather than becoming a hindrance. The Office of the Assistant Director 
lent extensive help to a study, commissioned by the Executive Officer, PHS, 
that resulted in a major PHS decision to consolidate all Parklawn ADP 
facilities. Under the guidance of the Office of Assistant Director, a 
systems study and a draft RFP were developed to allow an upgrade of DCRT's 
PDP-10 image Processing Computing Facility. Similar efforts were started 
regarding the main DCRT operated NIH central facility. 

As regards DCRT's own operation, the Office of the Assistant Director 
coordinates all ADP procurements, contracts and work orders which have a 
technical or engineering impact on our computer operations. By bringing 
an understanding of technical considerations into these matters, better and 
more responsive services are obtained from the central service groups. 
During FY74 the initial phases of two major physical site changes were 
coordinated. One is an addition to Building 12 to provide storage space 
and relocate office space. The second will be a gradual conversion of the 
second floor of Building 12 to machine space. 



12 



July 1, 1973 thru June 30, 197»* 

PUBLIC HEALTH SERVICE-NATIONAL fNSTlTUTES OF HEALTH 
DIVISION OF COMPUTER RESEARCH AND TECHNOLOGY 

1. DCRT-2 



Serial Number 
2. COMPUTER CENTER BRANCH 3. J. D. Naughton 

Branch Chief 



MISSION 

The Computer Center Branch designs, implements and operates a 
large general-purpose computer utility to meet most effectively 
the dynamic and diverse requirements of both N.I.H. research 
investigators and managers In the support of modern medicine. 
Tills charge Includes the original development of new system 
facilities to meet the unique requirements of the NIH mission in 
order to bring the full power of the computer to bear on 
problems at every level of biomedical research in many diverse 
locations. The core of this computer utility is a network of 
computers and remotely located terminals, which, by means of a 
modern communications network, extends the power of the computer 
directly Into research laboratories and administration offices 
throughout N.I.H. This provides immediate access to the 
computer from thus minimizing delays In the research program and 
making more efficent use of critical manpower than more 
traditional methods. An inherent responsibility of the Computer 
Center is the continued research and development of new methods 
to extend the network even further into the research environment 
while continually adapting to the constant impact of new 
knowledge and program direction. 

A full spectrum of computational services is provided to all 
Institutes and Divisions of the NIH on a fee-for-serv ice (cost 
recovery) bases. These facilities include conversational 
programming, graphics, microfilm output, text editing, remote 
job entry, tir.ie sharing, data base management and batch 
processing. Large systems as well ds mini-computers and 
terminals a.*e tied together providing a "distributed capaci»:y" 
available at many levels. Research into the computer and 
information sciences coupled with close cooperation between the 
N.I.H, medfcal Invest Sgators and the computer scientist have 
introduced computers directly into the research environment 
where they can perform most effectively in attacking the complex 
problems of medical research. 

The medical research programs of N.I.H, require the most 
powerful ard flexible of computer services and tools available 
today. The computer network provided must have a distributive 
power that Is easily accessible when needed to scientists in the 
laboratory Itself. The goal is to mold, polish and, in general, 

13 



enhance the computer Into a complete tool for medical research 
and its administrative support. New areas of computer 
applications are sought out continuously in conjunction with a 
comprehensive training program to Inform research investigators 
of the latest methods in the use cf computers to most 
effectively meet the unique requirements of their Individual 
laboratories. 

197U ACTIVITIES 

Throughout 1973, users of the NIH Computer Center .continued to 
demonstrate outstanding innovation In the application of all 
computational services to the mission of the agency. Users 
rapidly adapted existing programs to talce maximum advantage of 
new time-sharing services as they became available and continued 
to develop a never ending stream of new programs using 
traditional batch processing services. 

The overall workload of the Center continued to increase at its 
normal growth rate of 25% per year for the seventh consecutive 
year. The use of batch processing services increased throughout 
the year to a workload of 3,200 jobs per day with over 80^ of 
them being completed with results available to the user In less 
than two hours. 

The use of interactive terminal systems (WYLBUR, TSO, and CPS) 
experienced the most dramatic Increase (33|) to over 2,500 
sessions per day, A new record for the number of simultaneous 
UYLBUR users was set at 189 In April while the new TSO service 
established its first plateau of 36 simultaneous users. As the 
number of keyboard terminals using the NIH Computer Utility grew 
to 800, the number of telephone lines accessible to the system 
was increased to over 250 to minimize the frequency of busy 
signals received when dialing the interactive services. 

The current workload, along with the projected increase In 
demand for services, has precipitated i» system study addressing 
the problems ot expanding the computational capactt/ of the 
systeT and improving response time for Interactive services. 
This study has shown a need for significantly increased CPU 
power within the mul ti -processor utility. Procurement 
procedures are now being followed so the Computer Center can 
acquire the necessary capacity before service to the NIH 
community degenerates to an unacceptable level. The study has 
also identified a need for virtual hardware and its associated 
software techniques In order to maintain rapid response for the 
Interactive terminal user. These techniques will also improve 
system efficiency and will eventually permit the relaxing of 
some technical constraints currently placed on users of the 
Utility. The study also identified the need for new 
telecorrmunicat ion controllers to provide users the ability to 



14 



communicate with the system using CRT terminals at speeds of 
1200 baud. The controller will also allow more efficient use of 
existing teleprocessing ports. A variety of terminals, 
operating at different speeds with different protocols, will all 
be able to call the same telephone number thus reducing the need 
for redundant system resources. 

Increased workload, larger systems, and internal improvements on 
operating efficiency combined to permit the most significant 
rate reductions ever announced for users of the NIH Computer 
Center. Beginning January 1, 197U, both batch processing and 
terminal elapsed (logon) time rates were reduced 25-30°^ for all 
users of the Center. On that date, elapsed terminal time for 
all interactive systems, WYLBUR (MILTEN), TSO, CPS, CiCS, and 
the DECsystem-10 were reduced 25^ to the new rate of $3.00 per 
hour. The batch processing rate experienced the most dramatic 
change. The basic machine unit rate for all classes of batch 
processing work was reduced a full 30^ resulting in a new rate 
of $1.70 per resource minute and the night processing 
differential was increased an additional 5% making It 25% 
cheaper to process work at night. Tape and disk mounting 
coefficients were adjusted resulting In an overall reduction of 
61 for private tape and disk users. 

These rate reductions were made possible primarily because of 
the constantly increasing demand for computational services from 
the many users of the fi I H Computer Center. As the MIH Computer 
Utility expanded both in size and function to meet this demand, 
the economies of scale contributed significantly to reducing 
operating costs. Since the fixed costs and overhead of the 
system do not Increase In direct proportion to Its processing 
capacity, the effect Is to reduce per unit costs even further. 
This represents the seventh consecutive year during which the 
iJIH Computer Center has been able to offer rate reductions to 
Its users. 

Continuous efforts by the Computer Center Staff to improve the 
internal operating efficiency of the NIH Computer Utility also 
contributed significantly to reducing operating costs. Monitors 
were used extensively to tune the system to the unique 
characteristics of the NIH workload. Channels,, disks, tapes, 
and unit record equipment were reconfigured to obtain maximum 
throughput. Older communications controllers were replaced by 
more modern units offering new functions as well as improved 
price/performance. Major parts of the operating system were 
modified extensively to gain Improvements 'n Internal 
performance as well as increased reliability and function. 

In December the transfer of all interactive data lines to a 
newly ihstalled "data switch" was begun In an effort to separate 
the data traffic from the overburdened NIH voice switchboard. 



15 



The system itself also received new communications controllers 
to provide increased accessab i 1 i ty to the system. During the 
year the number of batch RJE terminals entering the system 
increased substantially and their transmission rate was 
increased from if^SOO to 7,200 baud to allow higher speed remote 
printing. As of the close of the year, almost 93| of all worl< 
processed at the Center was received over the teleprocessing 
network. 

As users developed new on-line data base applications the demand 
for on-line disk space increased dramatically. Additional FILE 
packs, PDS packs, TMP packs, work space packs, and private packs 
were added to the system in an attempt to meet this need. Over 
7 billion bytes of directly addressable disk space contains over 
80,000 users datasets on-line at all times; a necessity for 
effective use of interactive systems. 

In an attempt to keep pace with the changing requirements of the 
NIH computer users, many past restrictions were eliminated or 
eased and new systems software facilities were implemented. The 
maximum allowable region size for all job classes was increased 
50% to 300K bytes to permit the development of larger 
applications programs without the necessity for segmentation or 
overlays. A new disk mapping facility was developed to provide 
users with more detailed information on the type and status of 
their on-line datasets. A new tape copy utility was developed 
to allow users to copy entire tapes using minimal JCL. 
Facilities were provided to enable users to determine the amount 
of available free space on the on-line packs at any instant In 
time. The requirement that all procedures to be catalogued be 
carried to the PAL Unit was eliminated with the announcement 
that such procedures could be submitted via WYLBUR. 

Users received Individual assistance from the Computer Center 
staff on specific problems through written replies to 2,10lt 
Programmer Trouble Reports (PTR) and through thousands of 
telephone calls ind visits to the PAL unit for personal 
assistance. In an effort to minimize paper work for buth thvij 
Center and users, and in an effort to provide more effective 
support to remotely located users, the PTR and Refund Request 
forms were combined Into one and a facilit/ for submitting the 
new form via TSO was offered- The requliement for special 
registration for CPS use was eliminated and CPS users were no 
longer restricted to the use of only one library. 

VJYLBUR was enhanced with the addition of many new and Improved 
commands. TSO moved from an experimental test status to a fu!l 
production system providing terminal I/O and interactive program 
development and debugging in FORTRAN, COBOL, PL/1, and Assembly 
Language. WYLSUR and CPS were then made available under the TSO 
monitor. Al ::hough TSO experienced some rather severe 



16 



performance problems when first Installed, Incremental 
Improvements were made as the year progressed and additional 
Improvement is planned for next year. 

A major system project was Initiated for the redesign of the 
popular conversational text-editor and remote job entry system, 
WYLBUR. The new design is based on the experience of the past 
five years. It will provide a more efficient system, having the 
ability to handle over 300 terminals simultaneously. riew 
features will be added to the system. Many of these new 
features were based upon suggestions made by users over the past 
few years. New terminal types v/ill be available at higher data 
rates and a powerful new macro-processing capability will be 
available. This will enhance the power of WYLBUR and enable it 
to operate efficiently In many new areas such as transaction 
oriented on-line data collection systems 

The installation of Release 21.6 In June was the only operating 
system release change made In over a year, and the conversion 
was totally transparent to most of the user community. Those 
users who experienced the traumatic experiences of early release 
changes recognize the significant improvement in this area. 
Several compilers were replaced with newer versions fixing many 
old problems and adding new features. The PL/1 Optimizing 
Compiler replaced PL/1(F); FORTRAN G was retired In favor of 
FORTRAN Gl and ANS COBOL II gave way to ANS COBOL IV 

One of the more significant accomplishments of the year was the 
design and Implementation of an automatic dataset migration 
facility to both save money and make more on-line soace 
available for users. During the first two months of Its 
operation over 20,000 datasets were migrated to magnetic tape 
saving $30,000 In storage charges and making 80,000 tracks of 
critically needed on-line space available for use. 

As ^varnlngs of Impending paper shortages were received, many NIH 
computer userf. converted programs to produce output on 
microfiche rather than the traditional and more costJy printed 
page. Because of an overwhelming increase of over 200? in the 
use of microfiche during the year, a second unit was Installed 
at year's end to support the Increasing d«-nand. A new software 
utility was provided to allow a programmef to conveniently 
direct the complete output of a job to microfiche. 

The technical documentation and training activities of the 
Center continued to expand. The training pro.'^rc^m offered 
thirty-eight different courses and seminars during the Spring 
semester and forty in the Fall. Two new series of courses were 
Included among the new offerings for the first time this year--a 
series on the use of TSO and an expanded DECsystem-10 
programming sequence. Fourteen seminars extended the program 



17 



more deeply into the substance of the users problems. Over 
1,9U0 student course applications were received this year from 
candidates of varying backgrounds, an Increase of 5-1/2% from 
last year. Unfortunately, due to lack of qualified teaching 
manpower and classroom space, over UOO stuJents had to be denied 
adml ttance. 

As the number of users and functional capability of the NIH 
Computer Utility Increased, the demand for current technical 
documentation Increased correspondingly. The Technical 
Information Office distributed over 32,000 . pieces of 
documentation, ranging from a one page WYLBUR reference card to 
a kl& page Computer Center Users Guide. Documentation profiles 
of 2,700 individual users were maintained by the Automatic 
Documentation Service to Insure that all users received only the 
documentation relevant to their interest or needs. Many 
documents were completely revised and others received only minor 
corrections and enhancements. 

The frequency of INTERFACE decreased slightly this year as only 
seven Issues were published. These Issues presented over 275 
pages of technical documentation on the organization, policy, 
philosophy and use of the tilH Computer Center as well as basic 
concepts in computer science. Data Line continued to receive a 
number of provocative questions from users and provided 
Informative answers (we hope). Programming Methods appeared 
regularly with discussions of varied programming techniques and 
concepts In computer science. The Bugs, Hints, and Diagnostics 
section continued to provide users the detailed technical 
Information necessary to process work on the Utility. With the 
last Issue of the year, the Compleat Computor expanded In 
purpose to Include tutorials on subjects of general Interest, 
The second annual Index to INTERFACE covering all Issues up 
through the end of 1973 was published. 

The PDP-IQ system, now designated as DECsystem-10, also 
underwent major transformations during the year. Early In the 
year the original KA-10 processor was replaced by a n?w paged 
dual processor KI-1077 processor with hardware double precision 
arithmetic, faster memories, triple the amount of on-line disk 
storage, fast swapping drums and additional peripherals. 

This dual -processor system now provides additional capacity as 
well as processor redundancy permitting the system to remain 
operational even If one of Its processors fails. The dual 
system Is controlled by a new operating monitor which Is capable 
of dynamically balancing the load between the two CPU's. The 
accessibility to the DECsystem-10 was increased by providing 
ports for both 27U1 type interactive keyboard terminals and full 
duplex CRTs operating at 1200 baud. 



18 



The excellent Time-Sharing services of this system have been 
recognized by many scientifically oriented users. In 
particular, the constantly Improving and evolving OMNIGRAPH 
system has proven to be of great and continuing Interest to an 
ever growing group of graphics aficionados. A program named 
"PLOTX" provides the link between pictures seen on a scope and 
the hardcopy desired for research notebooks or formal 
publ Icat ion . 

Uf particular interest to laboratories with their own smaller 
data acquisition systems was the design and Implementation of 
software to permit a remote PDP-11 to be used as a terminal 
concentrator over a high-speed 9600 baud non-conditioned line. 
This system will handle up to 8 teletype devices with speeds up 
to 2U0 characters per second and is connected to a PDP-11 
Synchronous Communications Processor which In turn Is interfaced 
to the dual KI-1077 system. 

The already extensive repertoire of available languages and 
programming tools was supplemented this year by 2 major 
additions: APL and SPSS. The former language Is particularly 
suited for statisticians dealing with problems involving 
matrices and the latter Is of Interest to researchers with 
statistical problems. For those people who have need for 
special pre-prlnted forms or high volume printed output, a new 
facility was developed which permits the transfer of magnetic 
tapes containing upper-lower case text to the 370's for printing 
without the necessity of writing JCL. 

The operating hours of the DECsystem-10 were expanded 
considerably during the year to provide more flexibility In 
working hours for users of the system. The system now has 
operator coverage 16 hours a day from Monday through Friday and 
8 hours each Saturday. Even Sunday Is now available by special 
request . 

We are particularly pleased that the DECsystem-10 helped to save 
the life of a six-year old boy who had ingested an unknov/n 
poi'iop. An EPA gas chromatograph-mass spectrometer was used to 
analyze the sample and the sample's spectrum was matched with 
the standard parathion profile using the system. T!ie results of 
this search provided the basis for selecting the proper 
treatment. 

The XRAY modeling system has become a routine part of the 
process of the determination of the structure and function of 
macro molecules. A number of papers, films, and seminars have 
resulted from the application of this system. Extensions to the 
system during the year included the capability cf fitting 
molecular structures to electron density maps, and a means of 
determining the conformation of structure In the active site of 
a macro molecuie using fJMR data. In conjuncion v/ith modeling. 



19 



the crystal structure data base continued to develop and grow. 
The structures In the crystal data base can now be searched on 
bibliographic and structural terms. The resulting three 
dimensional coordinates of the structures are then manipulated 
using the XRAY system and displayed on a CRT for immediate 
review by the investigator. 

The first phase of the Computer Center's planned efforts to 
protect the confidentiality of sensitive data, improve the 
physical security of the computer system, and prevent 
unauthorized use of account numbers became visible. In addition 
to the Installation of locks on the machine room doors, two new 
software facilities were installed to provide data security. A 
data encoding system which scrambles sensitive data using a code 
phrase which is defined and known only to the user was made 
operational for all users. Another degree of security was 
achieved by Improving the existing keyword facility and 
extending it to make the setting of a keyword effective for both 
batch and terminal systems. A number of additional improvements 
for security and confidentiality within the NIH Computer Utility 
are under development for installation In the coming year. 



;975 PL AN? 

The high standards of performance and the outstanding reputation 
maintained by the NIH research community make It mandatory that 
the Computer Center continue to provide new computational 
capabilities to the NIH research administrative communities. 
New facilities plus the current high level of computational 
support are necessary for the success of the NIH mission. In 
the coming year plans have been made to increase the 
computational capacity of the NIH Computer Utility. This will 
keep the level of service at the desired plateau and provide us 
with virtual hardware, which will enable the Center to employ 
new virtual techniques in overcoming computational restrictions 
currently placed on the NIH research community. A comprehensive 
data base management system will be inscall«?d and pwt into 
general use. This will give users a powerful tool for data base 
management and rapid response to queries for specific 
Information. An entirely new WYLBUR with mary new and poverf'il 
features wll! be offered. The size of the on-line data storage 
will be increased significantly to meet the on-line space 
requirements of new applications. New terminal types will be 
investigated and the ones most appropriate for the NIH need will 
be offered to users. Programs for the corre'ation between 
crystal lographlc structure determination methods and NMR 
structure determination methods will be investigated, l/ork will 
continue In the fine tcning of the NIH Computer Utility to keep 
it responsive to the changing needs of NIH, and in the 
investigation cf new computational techniques that will 
contribute most effectively to the support of the NIH mission. 



20 



July 1, 1973 through June 30, 1974 



NATIONAL INSTITUTES OF HEALTH 
DR/'ISION OF COMPUTER RESEARCH Mm TECHNOLOGY 

Summary o£ Branch Activities 1. DCRT 



2. LABORATORY OF APPLIED STUDIES 3, Eugene K. Harris 

Chief 

The Laboratory of Applied Studies engages primarily in cooperative studies with 
NIH scientists, applying mathematics, statistics and computing science to bio- 
medical problems. In addition, the Laboratory undertakes mathematical and 
statistical research with applications to biology. 

During most of this reporting period, the Chief, LAS, has been on research 
assignment to the Medical Research Council of Great Britain, Clinical Research 
Centre, near London. The term of this assignment to the U. K. will conclude 
in August, 1974. Results will be forthcoming during the next reporting year 
and will be described in the next annual report. 

The Head of the Biomathematics and Statistics Section has served as Acting 
Chief. 

During this reporting year, the Medical Applications Section has completed 
evaluations of three conputer programs for the interpretation of electrocardio- 
grams. These results will appear in the July issue of Circulation . All routine 
adult electrocardiograms collected at the Clinical Center are currently being 
processed by one of the programs. After review by the cardiologist, the com- 
puter printou.t becomes the official ECG reading and is entered into the patients' 
charts . 

Evaluation of ECC programs is continuing, and additional cases with clinical 
documentation are being collected by the Medical Cardiology Department of the 
Royal Glasgow Infirmary in Scotland. Performance of a program in routine use 
at NIH will be comaared with that of a program developed in Glajgow. ECG data 
from Scotland, The Netherlands, and Sweden will be used in further testing and 
evaluation of programs . 

Conputer-based support of diagnostic and research activities of the Nucleai 
Medicine Branch of the Clinical Center was continued by the nuclear medicine 
task force consisting of personnel from both LAS and CSL of DCRT and those of 
MVI, CC. The most riportant clinical application achieved in FY 74 is the 
development of ECG-gated scintiphotographic cine angiograms of the heart. [ 
This tool makes possible the ncn-invasive study of heart functiOi''. in patients 
about to undei'go cardiovascular surgery and in a follow-up period. Becai'se of 
its non-invasive, non-traumatic character, requiring no anaesthesia and less 
radiation dose to the patient than a single chest X-ray, this study can safely 
be repeated frequently. Therefore, it shows considerable promise as a clinical 
tool. 



21 



Several other clinical applications including a scintigraphic evaluation of pul- 
monary function in pre-operative patients at NHLI are being studied. 

Activities in the conputer modeling of neurophysiological phenomena continued. 
These center around the LAS MAC- 16 laboratory coiipu,er. A neural hardware 
model interfaced with the MAC-16 was used to study a small neural net and give 
insight into the manner in which an action potential can modify the shape and 
duration of post-synaptic potentials and their spatio-temporal interactions. 
The system was also used to model cerebellar function. Study of the reciprocal 
inhibition-excitation behavior of small neural nets continues. 

Development of the MAC-16 laboratory research system continued. It has now 
been interfaced with the Marquette tape drive, losed for analog recording of 
ECG's, the Honeywell 7600 analog tape transport, the neural control panel, 
the general purpose switch- filter network, and a real time spectrum analyzer'" 
with ensemble averager. With this system, the investigator can autom.atically 
pre-process (filter and/or digitize) -- his data so that he can get the most 
economic use of a large-scale digital coirputer. 

In FY 74, the advantages of such pre-processing was demonstrated with cardio- 
vascular data recorded on live monkeys, which enabled the investigators to 
study myocardial contractility and ECG reproducibility. 

In collaboration with the National Institute of Occupational Safety and Health 
(NIOSH) analog data from vibratory equipment and the effect of this vibration 
on human electromyograms is being studied using the spectrum analyzer. 

The implied Mathematics Section has concentrated its efforts in two primary 
areas: (1) the mathematical modeling of biological processes, and (2) methodo- 
logy and techniques for fitting models to data. The various aspects of these 
areas are described in detail in the following paragraphs. 

Preliminary results from Oxygen Transport to Tissue Modeling have been presented 
at conferences and an extensive paper detailing the mathematical model has been 
submitted for publication. A nimber of simulation programs have been combined 
into a single program v/ith options for a variety of physiological conditions. 
Further work in this area will be concerned with inproved data on oxygen-hemo- 
globin dissociation reactions, autoregulatory effects of flow aiid tissue meta- 
bolism, and forj^alation of new concepts using time dependent flow fields. 

An extensive investigation of the partition of long-chain fatty acids between 
n- heptane and an aqueous buffer has been completed, and the results are in 
press. Related studies continue. 

The curve-fitting routines of MODELAIDE were revised, and a collaborative 
project with CCB and PSL was undertaken to adapt MODEIAIDE for use from a 
Tektronix 4010 graphics terminal under the new Time Sharing Option (TSO) . 
This system is now operational. 

A number of algorithms offering alternatives to least squares curve-fitting 
have been developed on a trial basis. 



22 



Ongoing mathematical modeling of the passive integrative behavior of single 
neurons has considered the transient response to individual synaptic inputs at 
single branches in the dendritic tree. This work, now in manuscript, has been 
presented at the American Physiological Society's Colloquium on Membranes, Ions, 
and Impulses during the 1974 FASEB Meeting. Present research efforts on active 
nerve conduction continue to exploit a siinple FitzHugh-Nagumo equation to study 
the dynamic stability properties of its traveling pulse solutions and their 
relation to the stimulus -response characteristics of this model axon. 

The activities of the Biomathematics and Statistics Section continued along the 
lines previously established. Study of statistical distribution theory appro- 
priate for multivariable measurements, and particularly for the multivariate 
lognormal model, was continued. These studies extend previous studies of size 
and shape variables and treat covariance properties of the multivariate log- 
normal distribution. Another line of research is that of population models. 
Discrete time models were studied, and a graphic "increment" plot is being 
studied as a practical tool for discrimination among models . Tfiese latter 
studies are undertaken in conjunction with a credited research project of a 
STRIDE trainee at American University. Several manuscripts using discrete time 
models in conjunction with conputer simulation have been completed along with 
a manuscript on the lognormal distribution. 

Also inplemented was a conversational con^DUter system which enables non-pro- 
grammers to transfer sequential datasets between conputer disk storage and 
private magnetic tapes, while maintaining a directory of the tape contents. 
It is now in use by personnel of the Laboratory of General and Comparative 
Biochemistry, NIMH, for scientific manuscript storage and retrieval. 

Work continued in mathematical studies related to linear algebra, the theory of 
graphs and linguistic analysis, and the theory of computability. Several publi- 
cations resulted from this work. 

Collaborative effort continued on an NCI project for conputer analysis of histo- 
logical and cytological bladder cancer pictures. An expandable interactive 
conputer software system has been ijnplemented and is described in the EMB report, 

Two members of the laboratory were appointed to an adhoc committee for the 
Federal Working Group on Pest Management, an interagency group. System models 
for the transport of DDT and lead through the environment were evaluated. 



23 



LAS reports published or "in press" during FY 74 : 

1. Bailey, J. J., Itscoitz, S. B. , Hirshfeld, J. W, , Jr., Grauer, L. E., and 
Horton, M. R. : A method for evaluating computer programs for electrocardio- 
graphic interpretation. I. Application to the experimental IM program of 
1971. (To appear in Circulation , July 1974). 

2. Bailey, J. J., Itscoitz, S. B., Grauer, L. E., Hirshfeld, J. W. , Jr., and 
Horton, M. R. : A method for evaluating computer programs for electrocardio- 
graphic interpretation, II. Application to version D of the PHS program 
and the Mayo Clinic program of 1968. (To appear in Circulation , July 1974). 

3. Bailey, J. J., Horton, M. R. , and Itscoitz, S. B.: A method for evaluating 
conputer programs for electrocardiographic interpretation. III. Reproduci- 
bility testing and the sources of program errors. (To appear in Circulation, 
July 1974) . 

4. Fletcher, J. E.: A mathematical model of the unsteady transport of oxygen 
to tissues in the microcirculation. In Oxygen Transport to Tissue , eds., 
D. F. Bruley and H. I. Bicher, Plenum Press^ New York, 1973, pp. 819-825.* 

5. Fletcher, J. E., Ashbrook, J. D. , and Specter, A. A.: Computer analysis of 
drug-protein binding data. Annals of the New York Academy of Sciences , 
Vol. 226, November 1973, pp. 69-81.* 

6. Green, M. V., Ostrow, H. G., Douglas, M. A., Scott, R. N., Myers, R. , 
Bailey, J. J., and Johnston, G. S.: ECG-gated scintigraphic imaging tech- 
nique for studying ventricular function. 21st Annual Meeting of the 
Society of Nuclear Medicine, San Diego, June 11-14, 1974. 

7. Green, M. V., Ostrow, H. G., Douglas, M. A., Bailey, J. J., and Johnston, 
G. S.: Scintigraphic cineangiography of the heart. MEDINFO Conference, 
August 5-10, 1974, Stockholm, Sweden. 

8. Harris, E. K. : Comments on statistical methods for analyzing biological 
rhythms. Chronobiology , L. E. Scheving, F. Halberg, and J. E. Pauly, eds., 
IGAKU SHOIN LID., Tokyo, Japan, 1974, pp. 757-760. 

9. Hutchinson, G.: Recursively unsolvable ;\ ird problems of modular lattices 
and diagram-chasing. Journal of A l gebra 26 1973, pp. 385-399.* 

.0. Hutchinson, G.: On classes of lattices represencaLle by modules. Proj^ 
ceedings of the University of Plousto n Lattice Theor'/ Conference , 197T,~ 
pp. 69-94 . * 

.1. Hutchinson, G.: The rapresentation of lattices by modules. Bul l. A mer. 
Math. Soc. , 79, 1973, pp. 172-176. 

.2. Kramer, R. J., Roberts, \. J., Gelfand, M. J., Milder, M. S., Hall, C. A., 
Frankel, R. S., and Johnston, G. S.; Ventilation-perfusion scintiphoto- 
graphy in eval.iating the pre-operative cardiac patient. (Abstract), 
Circulation 48: Suppl. IV-187, October 1973. 

24 



13. Marcus, M. L. , Shuette, W, H. , Whitehouse, W. C, Bailey, J. J., Douglas, 
M. A., and Glancy, D, L. : Use of a video system in the study of ventri- 
cular function in man. American Journal of Cardiology 32: 175, August 
1973. 

14. Mirvis, D. M. , Koph, F. S., and Pottala, E. W. : Off-line computer system 
for small laboratories. Proceedings of the Joint Meeting of the Profes- 
sional Association of USPHS , 1973. 

15. Mirvis, D. M. , Koph, G. S., and Pottala, E. W. : The effect of altered 
hemodynamic state on the harmonic content of the left ventricular pressure 
waveform. Proceedings of the 26th ACEMB , Vol. 15, p. 171, Alliance for 
Engineering in Medicine and Biology, Arlington, Virginia, October, 1973, 
(Summary) . 

16. Mortimer, J. A., and Pottala, E. W. : Possible sources of preferred centri- 
petal conduction in dendritic spikes in alligator Purkinje cells: a com- 
partmental neuron model. Proceedings of the Society for Neuroscience 
Annual Meeting (Abstract), November 1973. 

17. Pottala, E. W. : A hybrid neural siiraalation system. Proceedings of the 
1973 Summer Conputer Simulation Conference , p. 797, Simulations Councils, 
Inc., La Jol la, California, July 1973. 

18. Prewitt, J. M. S., Reece, D. K. , Hutchinson, G., and Jackson, C. K. : 
DECIDE: An expandable system for medical decision-making. Collogues IRIA , 
Informatique Medical, Le Chesnay, France, 1974, pp. 153-187^ 

19. Rinzel, J. M.: Transients in dendritic trees. (To appear in Proceedings 
of the FAi^EB , April 1974). 

20. Si]T5)son, R. B. , Ashbrook, J. D. , Santos, E. C, and Specter, A. A.: Parti- 
tion of fatty acids. (To appear in Journal of Lipid Research , July 1974). 

21. Specter, A, A., Santos, E. C, Ashbrook, J. D., and Fletcher, J. E. : 
Influence of free fatty acid concentration on drug binding to plasma 
albumin. Annals of the New York Academy of Sciences , Vol. 226, November 
1973, pp. 247-258.* 



*reported as "in press" in FY 73. 



25 



Corputpr Gystprs Laborrtory /^nnurl Report, IHTU 

Trblp of Tontpnts 
I ptroi^Lict i on 

I. Contipuinr Corrutpr Support for Pp-lipp Lpboratory 
Txppr i rentPt i op 

II. Ivio *'pv; l.pl'orrtory Conputer Projorts 

III. Clipicpl Carp rnr" Ppsparoh 

IV. ("pnorrl Resparch 

V. Gppcial Ponsultation 

Tatip of fanpovv'pr an<^ Expppditurp by Projprt 



27 



I nt ror^uct I nn 

Thp prlr?ry nissior of thr Corputrr Systens Laboratory 
(CSI.) Is to ii-'pntify proMpp rrp?s ir liioppr^ic?! rpsp?rrh 
anc' clinical ce\re in which thr corputrr offrrs a potPntial 
for inprovpd rcsparch productivity or ir^provrd hpalth c^rr. 
Thp concentration of v;ork is on applications v'hprr rral-ti^p 
data collection, analysis, display, and rxpprinpnt control 
arr required, wherp economic cons i dp ra t i cps favor a srall 
computer or where equiprent proxiriity is irportant. 

The staff of C5L has, in addition to exoertise in both 
the enri neer 1 nr and prorranrinr aspects of laboratory 
conputinr and autoration, extensive experience ir v/orVinr on 
problems in the biomedical area. f'any of the laboratory's 
projects require a coordinated effort bptvpen enrlneers and 
computer scientists from CS! and researchers from other 
rivlslons and Institutes. 



28 



I. Continuin.p- Corputrr Support for Cn-lirp L?hor?tory 
Expp r I rpntrt i on 

C5L is pnr.'^rpc' in Fi nunl^pr of prrjpcts supportirr 
rpsPc-'rch in thp physical anr* Ipfipvorial scipncps. Thp 
prinpry poal of tfi i s support is ?utor?tion of thp rollpctiop 
ppH procpssinr of lahoratory <^rtP . 

This Het?, v;hirh ori"-inptPS in ? vari'^ty of forrs, fror 
e iiuman pushinr a t^utton in rpsponsp to p stiruius to pn 
analytiral instrunpnt npasurin.f' a physical propprty of a 
molpculp, nust first ho convprtec^ to a frrr v/h i rh can tr 
transni ttpd to a rop-putpr. Thp CGL pnri pppr i nr staf'^ has 
Hpvploppd such intprfacinp equiprpnt for a nurhpr of 
applications. CGL has dpvploppd tv.'o lal oratory conputpr 
systpns in f'lAfTT anr* onp in NIAIP v.'hich a i r^ rpsparch on thp 
structuro of coirplpx p'oIpcuIps such as protpins and Rf,','' . 
The analytical instrunpnts v/hlch TGI has intprfacpd to 
computers includes sppct ropho tomptprs, a nurlpar narpptic 
resonance (fJfT) sppct roretpr, a spectropol ar i rp ter , and an 
anino acid analyzpr, all of vh i ch are uspd to rrpsurp 
different physical propertips of poIpcuIps. ^ systpn 
developed by CSl. for IJIf'l' collects and processes data fror 
human subjects pa rticipatlnp in psych clerical 
experimentation. The information collected includes 
phy s I olori cal measurements and subject responses. The 
.p-eneral mode of operation is for a subject to respond to a 
computer controlled stimulus while the computer pvaluatps 
and rpcords thp suLjPct's rpsponses, reaction timp, and 
physi olopi cal statp. 

Cppprally, the smaller laboratory computer systpms arc 
devoted to a single user at a time, which is the . ase v/i •:b 
thp r'f'R system in NIAfTP, v^herc the computer is used for 
Fourier transform spectroscopy calculations. The f) I /" I r 
system, on the other hand, is cVsi rned to service a numbpr 
of experir^ents in one laboratory at the same time. Mere the 
data are beinr collectpd as part of research into the 
structure of i mmunol ori cal ly important proteins. The fJIt'!' 
and the larrer MIAtTP systems have been developed to 
communicate simultaneously with a numlpr of 1 al ora to r i es 
located rpmote from the cor^puter. Purinr the past ypar a 
mi crosppct rophotometer was interfaced to the flivrr computer 
to riake a total of five analytical instruments pnr' two small 
computers currently on-line. Rather than beinf' interfaced 
to specific instruments the f.'If'll system provides reneral 
purpose stations which permit the psycho 1 of'i ca 1 rcspprcher 
to connr?ct a selection of equipment specific to his 

29 



.iffprrpt ro„f,,.urPt.ons nrr PV.-^il.^Mr to thn rrsr^P r rhr r s . 



1 1 . 



Tv.'o Mpv.' Laliorr'tory Corpiitpr Projects 

i^SL has un<-'prt?kpn tv/o ppw labor?tory c 
in thp past yerr, opp in support of rpsp?r 
arr* arothpr invnlvinp- a hpv/ ipstruppnt 
rrsparch for soparatipr cpIIs. 



laboratory corputpr projpcts 
(ispH in f-pncpr 



Laboratory of Visiop Ppsparch 

A rpsparch laboratory has brer orrapizpr! ip thp 
Laboratory of Visiop Tpsparch, •J'"!/ ^or thp stiK'y of hov/ 
visual pirrrpts sprvp to ipitiate thp pho totrapsi^uct i op 
procpss IP thr rptipa apH hov; this process is lirkpr* to thp 
trapsrissiop of si.fpals fron thp visual pi rnpp t-copta i ni pt 
photo-pxci tpH rPFiop of thp cell to its remote prr-sypaptic 
terriipal. A varipty of physical ape' chemical pxperimpptal 
rpthoc^s will bp employpr' apH ip post cases thp Hata fror thp 
exper i r-ieptal laboratory ipstrurepts will be ip the forn- of 
apalor vol tare sip-pals which vary as a functiop of tire. 

- hardv.'ireH sifpal PVPrpPC'r api' corputer have beep 
lipker* to capture, store aPc* apalyze data fror these 
exper irepts. PresePtly software is bejp.p- v/rlttep which will 
cause various trapsfers apH rani pu Tat ions of Hata to take 
plarp. A sirplp kpyboar'^ interrorated by the corputer will 
be usei! to cormunicate the fupctiops Hpsirer' by the 
researcher \;hep the syster Is ip use ip the laboratory. 



Fluorrscepce Activated Pell Separator 



(FA 

VI h i 
c-r,c' 
s ir 
?pc' 
las 
thr 

row 

of 

fVf 

one 

det 

flu 

vo 1 

1 ir 



J\! n 
rs) 
cb, v;e 

Lo 

i 1 ar i 

i pco 

pr be 

f 1 
PS trr 
uni ^o 
1 pcte 
1 \r 
err^i p 
oresc 
ure t 
fit sc 



typ 
^re 
re (' 
s 

t i es 
r[ or 
an . 
uorr 
r'r^ o 

f^ i p 
ht s 
p ce 
ppce 
o He 
atte 



es o 
nov; 
PVF 1 o 
A 1 Pnr\ 
. Cp 
a te(! 
HpI 1 

seep t 

f thp 
rops . 
to SP 
catte 
1 1 t' 

at 
termi 
r d i s 



Fl 
ope 

pec! 

s 

lis 

i p to 

s pa 

1 

las 

Th 

para 

r s i 

efle 
tv;o 

ne 

tr il 



uoresc 
ration 
upHpr 
Sci ppt 
are re 

a spa 
s s seq 
i t-lit 
er, tb 
ese r*r 
te COP 
rPPl r. 
ct i OP. 

\.'avel 
cell 
ut i op 



epce 
al i r 
copt rp 
i f i c 
pHp red 
11 1 iq 
u p P t i a 
produ 
e f 1 u i 
ops pr 
te i per 
re use 
The 
ePFths 
def 1 ec 
is pre 



A c t i v 

the 
ct I y 
Labo 
diff 
u i d s 
1 1y t 
ces 
c' is 
e cha 
s . 
d i p 

Los 
/ lir 
t ion. 
septe 



ated 

rjri . 

Stap 
ra to r i 
crept i 
trean 
h rourh 
e 1 ect 
b rol-'en 
rrp<' a 
Ppe ^ 
the Gt 

A Ian 
ht sea 

A 
d u S i P 



Cel 1 
The tv 
for'' 
es, h 
ally f 
i 1 luri 
the 
r i ca 1 
i nto 
ppropr 
1 uores 
anf ord 
rs sy 
tter a 
f 1 uorp 
r a CO 



Sepa 

sy 
lini v 
ave 

1 uor 
rate 

I ea 
'i i 



I a te 
cpr\r 
sys 
stem 
pd C 

SC'=P 

pven 



ra tors 

s ters, 

ers i ty 

bas i r 

PSCPP t 

d by a 
r- anr-' 
rp? 1 s . 
spr i PS 
1 y pnr^ 
p and 
ter to 
uses 
oul ter 
ce or 
tlopal 



31 



piilr.p hri.'^ht nnnlyr-cr (P!!/^) rr, .-^ nrpps i^f pnply-.i^ in l-ofh 

systf^MS. r<it."> rnllrrtf^'' iisinr' tlif^ I'H/^ rnn !r forr-.-i r tr-' tn 

slio\/ tlir f' i s t r i I u t i nr> oT Intli riuorrsrrprr ■"■it' srpttrr 
vrrsiis niirilrr nf rrlls. 



rr.i hrs 


prop 


osf^i p sr.\ 


('pt'^ ft'ni" 


FACT, 


rni- 1 i m 


p 1 nyl ark o f 


pp 


ppp lor 


i ns t rumrnt 


to tl"^ ror'piit 


nion i tor -'n t n 


rol 1 


p r t i o p p p - 


tlir ropipiiLrr 


CPP 


p i thor lip 


for Iptrr rv 


pliiPt 


i OP, or i 


C'l X r.'i r 


1 ' p p n r 


' 1 r^^ t r 1 x < 


r(- 1 1 roup ts . 


Thi 


s is thr 


r,T[i.il' 1 p r)f 


luii I 


<l i p.r . W' 


( ' p t p op p s p r r i f i 


r rlippprl 


ro 1 1 pr tP(' '' 


p t P 


p n ( ' s p V p 


poss i 1 lo v/i til tlip I'M/* p 


[irov i ('ri' to 


f pri 1 i tPtr 


niO(' i r ! rpt i op 


<^^ 


(IptP -'I 


I prorporp t r 


tlir porii'Pl ror- 


siirii PS I p 


t r r r P 


t I n p p p ( ' 


s t P t i s t i r P 1 


P V P 1 1 1 


p t I o p of 


ph 1 p to prof' 


1 1 r p p 


liprf'ropy 


p pp|-i"ppppt 


f i 1p 


of ('istrl 



p 1 1 ror'pii 
p p ( ' f r (■ ■ f^ 

tprr. r:: 

rr p n ' ' \ ' i 1 
(' proppss 

srn t <' i ro 
t ppp \ r s 
of f 1 iioro''. 

SPT'P r-,1 1 
o\'pvrr, tl' 
, t"pl:p ror' 
rfsu 1 t s 

1 n p p . A 

1 1 1 '^ o I ', p r 
s t r i I n t i o p 
r 1 rnirp t o T 
porrp 1 I 7 
.'is triluitl 
of <M-.plp 
I'Uti OPS r(> 



tor 

tl;r lo 

I V'i 1 

1 ( ' p V o 

tlir <'p 

r t 1 y f> 

o r t r ' ' 

rrnrr 

r i X u 

p ron[i 

p p r i t^ o 

- pl 

.'ispl 

V p t i o p 

Tp 

rofipti 

PtI r'p 

OPS . 

y p ( ' r p 
r flirt 



ystri' t 
s ^ 1 pr'os 
1 i P to r 
lop rror 
LP. ^Pt 
III to ''i 
i r ipTf' i r t 
ppf' srpt 
hirb tl> 
u t r r cpn 

PS \'M 1 1" 

1 of \;t, 

p y 1 1 p i t 
, rorpp 
t P p P p 1 
t P t i fiPP 1 
of rii 

Tlir syst 
"" 1 1 1 f s ,T 
b r r r r f r 



n ro 1 1 pp t 
', vs trii I y 
f p r r t i 1 r 

r p I' ' s \- I i r 1 1 
p p p t r r i n f 
si s t o r p r r 
Ply i p to p 
tor vrrsMS 
r ri"^ is 
sort tills 
r^r rv i oi's 1 y 
i rl' is pot 
vi 11 I P 
r i s o p p p ( ' 
y s i s \ ; I 1 1 
f 1 1 n r t lops 
r V r s P p f ' 

rp' \'i 1 1 l-p 
(' i"P i p tP i p 
rrrrr. 



III. Clinic rl Caro anr' RFSPprrh 

Col. h?s, for thr Ir^st srvprri yr?rs, hrror-r 

inrrrpsinrly involvrH in ttir support of rrny of thr rlirirpl 
earn f'nr' rrsrprch functions Pt thp nil'. This b^s orrurrof' 
1 ocnusp of inrrppsinp unrirrs t^nr" i nr r^rx^ pyrrrppss L'V 

clinicians of thp potpntial of corputrrs anH autoration anr' 
r'ur to trrnrnc'ous incrcasps ir the ar^ount o^ f^ata that is 
analy2P<' anc^ PxapinpH for the r^iarnosis anr* trcatnpnt of 
patippts. Thr incrpasp in Hata lias rcsultPi' ir part fror 
a-^vancps in rpHical i ns trumpn ta t i on anr* tpchniqups vli i ch 
havp Ipr" to thp capability for ppnpratinr voluninous amounts 
of (^ata that in turn rpquirp soph i s t i ca tpr' ror^uction anr* 
analysis to hp of any usp. 

Thp prinary roals in applying autoration to clinical 
npf'icipp prv to havp corputprs pprforr functions vh i ch 
cannot he pcrforrprl ly ranual rpans anH to pprfor""' routicp 
clerical functions o^tan r^orp hy npf'ical pprsornp! so that 
thp talpnts, traininr, anri pxpcriencp of thpsp ppoplo can he 
norp pffpctivply uspf'. 

Thp projects Hpscrilcd hprc rro rclrtpc' to rany facets 
of patlpnt csrr anc* clinical rcsparch. The tpchnical 
r pqu i rpr^pnts oT all of thpsp projpcts ircluHp thp 
acquisition, storarc, analysis, anH (-"isplay of clinical 
r^ata, anr*, in rest cases, there is a nppr* for thpse 
functions to he perforrpH on-1inp anr' in rnal-tinp. 

Fu r til prr^ore, all rpquirp pxtpnsivp cooperation aron"- 
pnrineprs, conputPr scientists, 'pmrrrnrrrs, anr' r-'cr'ical 
personnp 1 . 

I!1!L I I ntpnsi vp Care Uni t 

Thp purposp of this activity is the continuous 
nonitorin<T of patients in the he-''rt surrery recovery area in 
orr'er to prrvit'e the earliest possil le 'etection of ;lnoi-ral 
or ''anrercus conc'itions. For over a year nov, a corpiiter 

systert has heen in opr^ration anr* pprforrinr this fiircticr 
for a sirrle patient ly collectinr anr' analyzic rcr, 
terperature, f Uj i >.' loss, ant' arterial anr' venous pressures. 
This capahility is to l-e extenc'er' so th.at thf sara functions 
can he proviripr' s iru 1 taneous ly for four patients (^y 

ni f'-s urine r, 1?7U. f^uture plans l^eyonH that tirn InduHr thp 
r'pve loprpnt of nothor's for thp autor^-itic control of I looH 
anr! ('rur infusion, with the control jiascc' upon thp analysis 
of noni to rpr' clinical sirnals. 



33 



f^linlcal Pptholory Lnt'orritory 

In FY73 C3L !^pr?in ?n pffort to ovrrcorr pro! Irris of 
rrlirhility, infr'rquritp flrxllillty, rnr' irsiifflcirnt 

caracity to h?nHlp thr Irhorntory vorlOoar' rrrspntot^ to thp 
rxistinr clinic?! 1 pbor? tor y rorputpr systpn. it v/?s 

i rrrr" i f te 1 y appprrnt t!^?t thr only rpplistic 'lopo- tPrr 
solution '-.'PS totrl rrplacpnpnt of thp rxistinr systrn. 
Moi'pvpr, cprtain of thp prollrrs nrrr'rc' innp'^iptr PttPntion 
Irforp thr lonror tPrr pffort ronl(^ Irrin. As a rrsult ? 
now tpst rrsult pntry systrri \ips r'rvrlorrt' for thr Chpristry 
^r.r' I'pra to 1 o'^y srrvirr. frst rrsults pre nov rntrrrc' vir 

CPT tpn^inrls rnr* sppcirl r'pvicps ('rsi<^nrd l^y CGI for thr 
acquisition of c'ata fron rciiltrr autor-ptir floor" rountrrs. 
Thr tprrinals ant^ sprcial intrrfarps pvf control Ipc' by tv/o 
r^i n i corputprs, orp in thp Hhrnistry srrvirr pnr' opp in 
ilpnpto lory, v/liich ptp in turn ronnrrtr'^ to Clinical 

Patliolory's err 3200 corputrr systpn for op-lipp i-'ata 
collpctirp. Thpsr sut^systms havp Ippp fully opr rational 
sinop latr surrrr, 1?73, provic'inp- significant irprovrrrnts 
in !)oth prrforrancp and rplialility ovrr thr systrr- that v.'as 
usprl prpviously. 

in ar'r'ition to thp alovp CGI also hrp-an. In rir'-107^/ 
thr autorv'tion of sorp of thr routinp operations of thr 
ficroiiolo.p-y Grrvicp, v/hrrr it v/as copsir'errr' npcpssary to 
invpstiratp nrthot'S an^* procrr^urrs hy v/h i ch such e'ata shoi'lr' 
hp fianr'lpc' prior to incorporating thrsr functions into a now 
clinical laboratory systrr^. This vyor'- is thrrrforr 

consi r'rred to br of rn int<^rir nature. Thr roal is to 

prrrMt tfir mtry anc' storap-p of tost rrnurst anr" r^-sult r^r^- 
in filrs on thr 300/370 systrn such that thr r^ata can ''■r 
latrr rrtri'^vr^' |y inquiries initiatP'"' fror a typrvritrr 
trrrinal in thr r'icrol iolory 1 al.'ora tory . Thr systrn is nov; 
opr'rpt'\Pf' v.'itb rrrarc' to trst rrqurst f^Plr, lut is not fully 
I rp 1 pi^i ntr-^' for tfst rrsults, s i n ~p autor-ftic r-rans for thr 
translation of rrsult f'ata to nactiinr rfP'''p\ 1p f o rr has not 
y r t I'r'pp f'rvi srf' . 

""GL irran pf forts to'-'arr" rrplacrnrrt of thr rntirr 
systrr:-' ly (-"ryr lop i np-, in cooppration \.'i th t!r qijpipp] 
Cpntrr, a Prqupst for Proposals (PFP) for a nrv; Clinical 
Laboratory Inforration Gystrn. Thr rr \irrr no corrrrcial 
systrr^s proposrc" v.'l; ich coul-' rrrt thr lopr tr r'^ rrnnirrnrnts 
anr' roals o'" tbr clinical laboratories ; tb rrrforr, CGI. h.''S 
unr'rrtakrn to r'rvf']op a nrv Cllplral balioratory Copputrr 
Gystrr'', a project rstinatc-^ to tabr alout thrrr yrars anr* 
rrqiii rr 15 to 20 !~"pn-yrars of rf^nrt. In the rran tinr, thp 



34 



Clinicrl CentPr has Hrtprnirrcl thpt tbo prrfi'^rrrncr of tho 
fxistinr systpr^ hps hrconp intolrr^l 1p, ppH ?> ror'rf-rr i r 1 
systn^ v/ i 1 1 bp oLtPinnr' for usr ?s r stoprpp n^ppsurr v/b i 1 r 
thp rSL drvrloppopt pffort is in prorrrss. 



''uclPrT f'rr'icinp PprPftnpnt 

CGL Iirs rontipupf' to rnrtrlfutP pprlrrprirr pxpprtisp 
to r joint projprt v/ith I/'^S, PrPT anrf thp Hurlpfr fprMrinp 
Ppppr trppt, nc to rorp fully r'rvplop thp potpntipl of ? 
rorputpr systpn v,'h i rli hps bopp r^ppr? t ionr>l for ?1nost two 
ypprs. Thp soirntifir oKjpctivPS ?nf' prrorrpl i sbrppts of 
this projprt aro r'psrribpH ip thp \..^Z rpport. Ip 

pprtirulrr, CGL bps partiripatpH in tlip r'rvplopnppt of 
tPcbpiqup for stur'yipr crrf^\ pr por f orr-pncp . This tPchnioup 
pprrits visualization of tho hpart rpvitips tbroiirb 
pop-invasivp procpc'uro. Both ciuan t i tp t i vp pnr' quplitativp 

ppplysis npthods rrc I.pinr i pvps t i ra tp*^ . 



Phonorari' I orrpr^ Hpsparrh 

For thp past tv/o ypars CGI. has rollaboratpH vi tb thp 
ourrical "ranch of f'I'l. I ip thp f'pvpl or)nppt of riptlioHs by 
\'/hich cba rartPr i s t i rs of ttip phonorarr"! orran^ ran Ip iispr' as 
Hiapnostic inHIcPS of prostl'Ptic hoart valvp pprforranro. 
Pbonora rf' i oFrar''5 pm nov/ routinply takpn on about six 
patippts ppr wppb. r>p(' analyzp^^ op tbp hytrir' oor^putor. Thp 
analysis ipvolvps bpat-to-boat rorrolation nf tho 

phopocar f' i or ran ani' a r'p to rr"i na t i op of thr ratio ^f tho 
anplituf'p of tbo opppipr soupH tn that of tbo rlosipr soupr'. 
Thoro is pv\ ''orcc tliat thpsp factors can bo usp'* to c'rtcct 
pc' iH^ntify cprtain prosthotic hpart valv?^ ra 1 f urio t iops . 



Pulr^onary '-'bysiolory l.aboratorv 

Puripf' FY7U, CGI. rorplptof' thp antoration of sopp of 
thp clinical aP"-' rosparch protocols oP thp Pulnopary 
Pliysiolop-y Laboratory, flM.!. TIipsp involvpH thp autoratiop 
of routipp pulrionf>ry funotiop pvaluatiops anr' of clinical 
rosparch artivitips rpquirinr tbp rpasiirppipn t of rospiratory 
rpsistancp. This pffort has bppp succpssful ip pliri'patipr 
thp pppf' for npr'ical tpcbpo lori s ts to porforr pxtppsivp 
npnual calculations and in allov/inr t!ip lal oratory to 
arcopodftP a significant ]rrrr'p':.r in v/orl'lord v/ithout 
corrrspoc-'inr- incrcasp in staff. 

35 



IV. Gpnrrnl Rpsp?rch ~ ' 

l.'li I 1 p thp bull: of thp v;orI' in CSI i? roprT^rtPr^ v/i th 
laborrtory autorrtion ?nH rllplrrl c^tp thprr is 

cons i ('prab 1 p pffort (-"pvotpr' to othpr rrp^s of rorputpr 
rrsprrch rplrtPf' to biorpdical rprlicatiops. Curroptly 
thprr arp two major arpas of o-pppral rp•sr^rrh^ tbr usp of 
ronputrr pattprn rrrornition nptbor^s in I ior'pHiral prot Ipr^s 
anr" thp (Vvplopnpnt of a npHical tP 1 pronrun i ra t i ops systpr^. 



Pattppp Rprornitiop Stur'irs 





l.'o rk con t i pupH 


in applyinp- pattprr 


tech 


niqups to stut'y 


thp rr 1 at i onsli i p !rt'. 


'ppp t 


and 


chpnical structurp 


of drurs. Thr rain projpc 


163 


anti-tunor drurs. 


thp chenical struct urr of 


dpscribpri nathpnaticall 


y !y couptinr tlip pu^^ 


^t pr o 


of chanical suhtructura 


1 unit it cop ta i npd . 


Two 


havp 


thp samp kind of 


suhstructural uni ts 


rrc 


a 1 i 1- 


p in structurp en 


d, assurinr that stn 


ic tu rp 


act i 


vi ty, they ?re al i k 


p in ac t i vi ty also. 


Thr a 


othpr f'rups of rolatpc' 


structure can then 1 


■p prp 


thel 


r structurp. Prrl 


ir-iinary rpsults i pd 


' i ca tP 


mpth 


oc's hold prori sp 


for scrprniPT ]^rrp 


nurhp 


v/h i c 


h may havp antl- 


turor act i vi ty . Th 


IP rrr 


?ppl 


icable to other s 


triicturp-act i vi ty stu 


1 r' i P S 


now 


[;rp-inninr- v/ith othpr data. 





rrcorn i t i op 
hp act i vi ty 

t I PVO 1 VPr^ 

\/h i ch v.prp 

f each typp 

druFS v;h i ch 

CO p s i (■■ p r p d 

rplatps to 

ct i vi ty of 

"■^ i c tpd f ror 

that thp 

rs of drurs 

thods arc 

an^^ \:or\' i s 



in othpr rrlatpd <,;ork brinr donp, chprical rorpounds 
arc charactpr i zpd by thpir rass spectra. T'pthods ?rc hpinr 
''pvploped to obtain thp rifht corbinatiop of r-assrs to 
drscrilp thp conpoijpds in nrc'er to ir^provp classification of 
upknov/n ccrpounds fron thpir nass sppctra. 



f'pc'ical Tr ^ eccrrn.n i cat i ons Gyston 

For spv?ral years C3I. has I^ppp en^Tpred in the 
('evelopnent of a technolory ly whicfi corputpr services can 
be nade available to thp red leal cor^runity us in"- oply a 
copventional touchto-ie tplephcne as a corputer terminal. 
Tfie plan has been to r-iake sucbi services readily available sc 
that the pov/er o^ thp computer can le applied in assistirr 
physicians in such areas as diagnosis, treatment, anr' 
therapy planninr. 



36 



A 
USP of 
provl (^ i 
i n vo 
appl i c? 
i nr"! c?t 
The np(' 
express 
?pp 1 i cs 
3 pro 
rol 1 ?bo 
venture 
ev? 1 uat 
project 



prototy 

the 
r\F i n p u 
I ce o 
t i on pr 
i ve of 
i c? 1 s 
er! i nt 
t i ons i 
up of 
rptp v;i 
i s exp 
Ion vj'i ] 



pe sy 

pushb 
t to 
ver 
or ran 
the t 
chool 
eres t 
n the 
rorr 
th th 
ecter^ 
1 he 



s tein h 
u t tons 
the CO 

the 

s have 

ypes o 

of 

i n 

Uni ve 
uni ty 
p Uni 

to CO 



as ! ' e 

on 
npute 
telep 
als 
^ app 
the 
u s i n ,p- 
rs I ty 
phys 
vers i 
n 1 1 nu 
as 



en corpi 
the te 
r , wh i 1 e 
hone, 
o been 
1 i ca t ion 
L! n i V e r s i 
the s 
hospi ta 
i c i ans . 
ty for 
e throur 
to the 



eteH 

1 eph 

tie 

A n 

s v/h 
ty 

ys te 

1, a 

CG 

thi 

h FY 

r'] 



v/h i ch 
one a 

corpu 
urher 
e 1 oper* 
i ch r i 

of i;i 

r-i f n r 

nr" pos 
I. has 
s pu r 



75, af 

spos i t 



vrrr\ t 

s a rrp 

ter res 

of ne 

v.'h i ch 
rht he 
s CO n s i r 

a se 
s i h 1 y 

prrrr 
pose. 
ter v/h i 
i on of 



s the 
ns o^^ 

r" i ca 1 

PTC 

user'. 

has 
t of 
anonp 
d to 
Thi s 
ch an 

the 



37 



V. Spprir'l Consult Ption 

In ?('rMtion to thr v/orl' r'psrrW'pd, CSL ronsults vith 
rPSPrTcfiers in nppd of corputpr pxprrtisp. This 

copsultPtion c?n !)p sirply ec'vire or ? sr^Tlfir ^ro^•lpr or 
cpn rosLilt in tlip (!psl<"n of r r'orp of sprrirl purposo 
hpr'r'\'Prp or in tlip '..'ritinr of spprial softwrro. 





' typicp 


nccurrr'' \j\t\ 


for \:] 


lich CGL 


v.'U i ch 


dp tprr- 


th?t < 


^^t^, in 


rsL 


p n r 1 n p p 


conriin i cat i on 


to tl 


IP P^P- 


prrcpss 1 nr. 


sho\'r 


thp fpp: 


othpr 


srell 1. 


to trcnsnit d, 



1 pxpnplp of til 
h p f'p ta CO 1 Ipct 
' s adyi cp v/as sou 
inpd t'np siZP sp 

(' i f i tf'i 1 f orr, on 
r r^odi f i pc' th 

i ntprf pcp so tha 
10 conputpr, vl 
The nodi -f" i cpt i ons 
s i h i 1 i ty of th i s 
aboratory .■'ata co 
ata for procpss i n 



i s typp o^ 
i on proh 1 pr 
pli t . ."> P i 
pctrur of pp 

a ra^^np t ! c 
is pqu i pnr 
t di p-i tal r^p 
a tPlp phono 

hui It ""or 
pppropch and 
1 Ipct ion sys 
r en thp cer 



support i vp 


v.'or!' 


a r i s i r p- in '' 1 A 


.rrr. 


nstrunpnt pxi 


StPd 


rticl 


Ips and storod 


tapp 


casspttP. 


A 


nt 


and pr'c^fd a 


ta can nov Ijp 


spnt 


1 i nps, for fur 


thpr 


this 


systpr 


havp 


can 


bp pr^pptP 


■d to 


tpns 


to al lov- 


ther 


trrl 


f a c i 1 i t y . 





38 



Summary of 

LEVEL OF EFFORT AND EXPENDITURE 
BY PROJECT 



Iject Name Project Leader 



DCRT Man-Power (M-Y/Year ) 
FY-73 FY-74 



DCRT^Capital Invested 
($X1Q00)1nc1uding maint. 
FY-73 FY-74 



Idg. 2 516 


Shapiro 


MR 


Col burn 


II MH 


Syed 


IIAID -Waxdal 


Shapiro 


lEI 


Schultz 


ICI 


Schultz 


.1 


Syed 


;iin. Path. 


Plexico 


Nuclear Med. 


Schultz 


■•honocardiogram 


Schultz 


iedical Tele- 


Plexico 


:.ommum" cations 




Pattern 


Shapiro 


Recognition 




'ulmonary Lab. 





.5 
.5 



2 

2 

2 

3 

1,5 

1 

7 

4 
.5 
.5 



6K 


2K 


20. 7K 


24. 8K 








5K 


9K 





IK 








80K 


80. 6K 


5K 





5K 


40K 


.5K 


.5K 


14. 5K 


3K 








2K 






39 



t 



July 1, 1973 through June 30, 1974 
PUBLIC HEALTH SERVICE - NATIONAL INSTITUTES OF HEALTH 
DIVISION OF CCM'UTER RESEARCH AND TECHNOLOGY 

Summary of Branch Activities 1 . DCRT - 5 

2. PHYSICAL SCIENCES LABORATORY 3. Dr. G. H. Weiss 

I . OBJECTIVES 

The Physical Sciences Laboratory is devoted to the study of problems in 
physics and chemistrv that relate to the biological sciences. Several 
disciplines are represented in the membership of the laboratory. These 
include applied mathematics, theoretical chemistry, and theoretical physics. 
Whenever possible the theoretical studies are performed in conjunction with 
experimental work, either in collaboration with workers in outside units, 
or by members of the Physical Sciences Laboratory working in other labora- 
tories at NIH. In addition to performing research of its own choosing, 
members of the Physical Sciences Laboratory provide consultation to other 
researchers at NIH on different topics in the disciplines represented in 
the Laboratory. These services are enumerated in the project reports. 

II. CURRENT LABORATORY PROGRAMS 

1. Research in. the general area of biochemical separation techniques has 
been carried forth with three objectives in mind: the elucidation of physical 
and chemical effects that bear on separation techniques; development of 
methods of processing data from analytical techniques, and development of 
new methodology in the separation methods. In this past year Dr. George 
Weiss has, in collaboration with Professor N. Catsimpoolas of M.I.T. and 
Dr. David Rodbard of NICHD, given a theoretical analysis of transient iso- 
electric focussing, a separation technique currently in u.se in the equilib- 
riuii regime only. Data from the transient phase can be us' i to furnish 
more detailed molecular parameters than has so far been possible. Experi- 
mental work on transient isoelectric focussing is being carried out by 
Professor Catsinqjoolas and his collaborators at MIT. Another project in 
this area is the elucidation of the effects o:^ iiDlecular heterogeneity on 
resolution in chromatography and electrophoresis. We have shown that hetero- 
geneity sets practical limits on the ability of systems to resolve molecular 
species. Further work is being carried out in the area of resolution in 
chromatographic ?ystems. The classical definition of resolution is inadequate 
for the design of systems with a gradient of properties. We are presently 
developing alternative methods for specifying the resolving power of a system 
that can be used for the more general systems now in lase, such as two di- 
mensional electrophoresis or pore gradient electrophoresis. 



41 



2. Another area of investigation is that of elucidating physical forces in 
biological entities in order to interpret the phenomenon of biological 
assembly (cell membrane formation, protein association, celliilar aggregation 
into tissues) in physical terms. Progress during the past year has been in 
the development of new expressions for interactiOii energies and the success- 
ful design of experimental systems to measure physical forces. The theory 
of van der Waals forces between large bodies has been developed to include 
the influence of spatially varying distributions of material at their sur- 
face. Drs. Parsegian and Weiss have also succeeded in formulating the 
attraction between parallel curved spherical and cylindrical bodies. A 
physical method has been found that facilitates confutation of electrostatic 
forces between rodlike particles, in collaboration with Professor S. Brenner 
of the University of Kentucky. The results have been applied to the pheno- 
menon of gel formation by Tobacco Mosaic Virus. A successful combination of 
experimental measurements by Dr. A. Shih at the National Bureau of Standards 
and theory developed by Dr. Parsegian has been used to analyze the attraction 
of single atoms or molecules passing in a beam near a solid substrate. This 
is an extremely sensitive technique for the measurement of forces on the 
molecular level. A second experimental method, for the measurement of 
forces between lipid membranes has been developed in collaboration with 
Professor P. Rand of Brock University, Ontario. It was found that the 
attraction forces between membranes vary with the presence of solute particles 
in the water as theoretically predicted by Dr. Parsegian. A method was de- 
veloped to exert forces on the membranes themselves to measure repulsive 
forces between them. 

3. Techniques of intensity fluctuation spectroscopy for measuring hydro- 
dynamic coefficients of biological macromolecules , swimming speed distribu- 
tions of motile microorganisms, and the electrophoretic mobilities of blood 
cells and platelets are being jointly developed by Dr. Ralph Nossal, Professor 
S.H. Chen of MIT, and Professor B. Berne of Columbia. Recent work on this 
project concerns the effects of optical heterogeneity on the inelastic light 
scattering spectra of large particles. The presence of chromatin rich regions 
within a cell leads to a reduction of the effective size of a microorganism 

as measured by its light scattering spectrum. 

4. Dr. Ralph Nossal in collaboration with Dr. L. Kohn, NIAMDD and Dr. B. 
Nisula, NICHD have developed refined Leukocyte Migration Inhibition (MIF) 
tests for detecting tissue antigens The assay can be adapted to test for 
cellular autoimmune bensitivity to thyroid tissues of patients hiaving various 
manifestations of thyroid disease. For exc'imple, it has been found that an a- 
granulocytic patient having Grau's disease show's a positive MIF response 
despite the absence of PMN leukocytes in their peripheral blood. 

5. Dr. James Ferretti in collaboration with Dr. E. Becker, NIAMDD, V. Colburn 
of DCRT, and T. Clem of DRS, has developed a fast scan technique to obtain 
NMR spectra. Two working NMR spectrom.eters have been modified to permit 
spectra to be scanned in very short times (often less than two seconds) . The 
responses are digitized, time averaged, and stored in a Raytheon 704 jonputer. 
The couputer is then used to crosscorrelate the data either with a suitable 
reference response or with an analytical function. The rapid scan technique 



42 



compares favorably with pulse methods in sensitivity and in some instances 
has distinct advantages over the pulse techniques. An important advantage 
is the sinplicity of the additional instrumentation required, 

6. Dr. George Weiss, together with Dr. D. Hoel, NIEHS, and Dr. R. Simon, 
NCI have continued development of adaptive saupling techjiiques to be applied 
to the design and analysis of clinical trials. Several methods were compared 
for choosing the better of two treatments in which success is measured in 
terms of survival time. One technique, suggested by Flehinger and Louis was 
found to be obviously superior and quite robust. This method was modified 

to give a stopping rule and adaptive sampling procedure for choosing the 

better of two dichotomous treatments (i.e., results are given either as 

success or failure). The resulting technique was found to be considerably 

more efficient than all others given in the literature. A further applica- 
tion was made to the problem of deciding whether a new treatment is better 

than a control treatment. Here too, the likelihood method proved to be the 
method of choice. 

7. Drs. Robert Jernigan and James Ferretti are studying the kinetics of the 
helix-coil transitions of polypeptides in solution. Relaxation times and 
time correlation functions have been calculated on the basis of a single 
helical region in the chain, as is usually the case for short polypeptide 
chains. The calculated values of these parameters can be conpared with ex- 
perimental results to obtain kinetic parameters. The results so obtained 
are m disagreement with values derived from the current theoretical inter- 
pretation, the differences being several orders of magnitude for short chains. 
Although the newer theory has not been validated experimentally (nor has the 
older one) it takes into account dependence on chain size, and so might be 
expected to be more accurate. 

8. Dr. James Ferretti, collaborating with Dr. N. Sharpless, NIAMDD, have 
been measuring NMR spectra at 220 MHz and analyzing the results using the 
coirputer program UEAITR written for the PDP-10. In this way they have found 
that the spin -coupling constants are not sinply related to the electron density 
but that the chemical shift and electron densities are linear] - related. 

9. Drs. Robert Jernigan and George Weiss have begun a study of the validity 
of tho rotational isomeric approximation for calculating configurational 
properties of macromolecules . For short chains it is possible to obtain, 
essentially by enumeration, effects of details of the energy contour on 
configurational properties. The distribution functions that are being calcu- 
lated allow one to check various approximations that are presently in wide use. 
Present results suggest that these approximations may lead to erroneous results, 
In connection with this project an international syTiiposium was held at NIH in 
February on Configurational Properties of Biopolymers. 

10. In addition to the individual research projects enumerated in the pre- 
ceding paragraphs, members of the Physical Sciences Laboratory engage in a 
large number of consulting activities in collaboration with ocher scientists 
at NIH as well as in the broader biomedical community. Specifically, Richard 
Shrager has made extensive use of curve fitting techniques on collaborative 

43 



problems. He has applied these techniques to fitting potentiometric and 
thermometric data for the titration of proteins supplied by Drs. R. Berger, 
NHLI and M. Marini, Northwestern U. . He has furnished technical assistance 
on the problem of the digestion of fibrogen by trypsin, with data from Drs. 
E. Mihalyi and D. Towne, NHLI, as well as on the determination of rate para- 
meters for fumarase kinetics with Dr. L. Kohn, MIAMDD and Dr. I. Darvey, 
IMiversity of Sydney. Mr. Shrager, D. Ashbrook, and G. Knott of DCRT have 
iiiproved and ipdated Modelaide and MLAB programs. Modelaide has been re- 
vised extensively and now runs on TSO. Mrs. Mildred McNeel has worked ex- 
tensively ^^dth Dr. W. Caveness, NINDS, and Dr. B. Rish, Naval Medical Center, 
on data relating to head injuries to veterans of the Korean and Vietnam wars. 
She and Dr. George Weiss have also worked with Dr. E. Fishmann of Freedmen's 
Hospital on the use of EKG surface maps as a diagnostic aid in left ventricu- 
lar hypertrophy and acute myocardial infarction. Mrs. McNeel has also help- 
ed Drs. Jack Cohen, NICHD and M. Hayes, NIAMDD to conpile a review on nuclear 
magnetic resonance of amino acids, peptides and proteins. Dr. Adrian Parsegian 
has led a tutorial seminar on The Physical Approach to Cell Membranes and Dr. 
George Weiss has given a series of lectures on Numerical Analysis for DCRT. 

11. Dr. Adrian Parsegian has been elected a member of the Council of the 
Biophysical Society. 

Dr. George Weiss has been appointed a member of the NIH Medical Board. 

PSL PUBLICATIONS 

Weiss, G. H. , Dishon, M.: Approximate solutions of chemical separation 
equations with diffusion. Advances in Chemistry 125 , 207-215 (1973). 

Rodbard, D., Weiss, G. H.: Mathematical theory of iinnunoradiometric 
assays. Analytical Biochemistry 52, 10-44 (1973). 

Johnson, M. , Yphantis, D. A., Weiss, G. H. : Instability in pressure- 
dependent sedimentation in monomer -polymer systems. Biopolymers 12, 2477- 
2490 (1973). — 

Weiss, G, H., Dishon. M. : Elution properties of inhomogeneous systems. 
Biopolymers 12_, 2631-2637 (1973). 

Weiss, G. H., Ackers, G. K. : Effects of nonumifom gel properties in 
analytical gel chromatography. Analytical Biochemistry 57_, 569-577 (1974). 

Rodbard, D. , Chrambach, A., Weiss, G. H.- Optijiazation of resolution in. 
analytical and preparative polyaciylamide gel electrophoresis. Electro- 
phoresis and Isoelectric Focussing in Polyacrylam ide Gel Electrophoresis, 
(to appear) . ' 

Weiss, G. H., Rodbard, D.: Resolution of species showing microheterogereity 
by zone electrophoresis and chromatographic systems. Separation Science 
(to appear) . 



44 



Weiss, G. H., Catsinpoolas, N., Rodbard, D, : Transient state isoelectric 
focussing. Archives o£ Biophysics and Biochemistry (to appear). 

Parsegian, V. A., Gingell, D. : A physical force model of biological 
interactions. Recent Advances in Adhesion (Gordon and Breach) 153-192 (1973). 

Parsegian, V. A.: Long-range forces in the biological milieux. Annual 
Review of Biophysics and Bioengineering, 2_, 220-255 (1973). 

Gingell, D., Parsegian, V. A.: Prediction of van der Waals interactions 
between plastics in water using the Lifshitz theory. Journal of Colloid and 
Interface Science 44, 456-463 (1973). 

Weiss, G. H., Kiefer, J. E., Parsegian, V. A.: Effects of dielectric 
inhomogeneity on the magnitude of van der Waals interactions. Journal of 
Colloid and Interface Science 45^, 615-625 (1973). 

Parsegian, V. A.: Possible modulation of reactions on the cell surface 
by changes in electrostatic potential acconpanying contact. Annals of the 
New York Academy of Science (to appear) . 

Nossal, R. J., Weiss, G. H.: Analysis of a densitometry assay for 
bacterial chemotaxis. Journal of Theoretical Biology 41, 143-148 (1973). 

Parsegian, V. A.: Formulae for the electrodynamic interaction of point 
particles with a substrate. Molecular Physics (to appear). 

Brenner, S. L., Parsegian, V. A.: A physical method for deriving the 
electrostatic interaction between rod-like polyions at all mutual angles. 
Biophysical Journal (to appear) . 

Parsegian, V. A., Weiss, G. H.: Electrodynamic interactions between 
curved parallel surfaces. Journal of Chemical Physics (to appear). 

Nossal, R. J., Chen, S.H.: Effects of chemoattractants on the motility 
of E. Coli bacteria. Nature, 244_, 253 (1973). 

Nossal, R. J., Weiss, G. H.: A generalized Pearson randun walk allowing 
for bias. Journal of Statistical Physics (to appear). 

Nossal, R. J., Weiss, G. ri.: A descriptive theory of cell migration on 
Surfaces. Journal of Theoretical Biology (to appear). 

Nossal, R. J., Berne, B. J.: Inelastic light scattering by large structur- 
al particles. Biophysical Journal (to appear). 

Jemigaii, R. L., Ferretti, J. A., Weiss, G. H,: Helix l:ifetimes within 
the conformational transition region. A random walk model. Macromolecules 
6, 684-687 (1973). 



45 



Ferretti, J, A., Jernigan, R. L.: Conformational lifetimes in the helix- 
random coil transition region by NMR. Macromolecules 6^, 687-69C (1973). 

Milstein, J. B., Ferretti, J. A.: The effect of polydispersity on the 
nuclear magnetic resonance of polypeptides. Biopolymers 12_, 2335-2345 (1973). 

Gupta, R. J., Ferretti, J. A., Becker, E. D. ; Rapid scan Fourier trans- 
form NMR spectroscopy. Journal of Magnetic Resonance (to appear). 

Sharpless, N. E., Bradley, R. B., Ferretti, J. A.: The nuclear magnetic 
resonance of heterocyclic compounds related to anthracene. Journal of 
Organic Magnetic Resonance 6^, 115-120 (1974). 

Hoel, D. G., Sobel, M., Weiss, G. H.: A survey of adaptive sanpling for 
clinical trials. Perspeccives in Biometry (to appear). 

Hoel, D. G., Sobel, M., Weiss, G . H. : Comparison of methods for choosing 
the best binomial population with delayed observations. Journal of Statis- 
tical Confutation and Simulation (to appear) . 

Kiefer, J. E., Weiss, G. H. : Truncated version of a play -the -winner rule 
for choosing the better of two binomial populations . Journal of the American 
Statistical Association (to appear) . 

Simon, R., Weiss, G. H., Hoel, D. G.: Sequential analysis of binomial 
clinical trials. Biometrika (to appear). 

Hoel, D. G., Weiss, G. H.: Conparison of methods for choosing the better 
of two negative exponential lifetime distributions in Reliability and 
Biometry (SIAM, Philadelphia) 563-583 (1974) . 

Dresser, A. Meirowsky, A. M. , Weiss, G. H., McNeel, M. L., Simon, G. A., 
Caveness, W. F.: Gainful employment following head injury, prognostic factors. 
Archives of Neurology 29, 111-116 (1973). 

Cohen, J., Hayes, M., McNeel, M. L.: Nuclear magnetic resonance of 
amino acids, peptides, and proteins. Magnetic Resonance Review (to appear). 

Wexss, G. H. : The diffusion constant for two-state brownons. Journal of 
Statistical Physics 8_, 221-224 (1973). 

McNeil, D. R., Weiss, G. H.: Merging from an isolated intersection. 
A dvances in Traffic Science (to appear) . 

Blumenfeld, D. E., Shrager, R, I., Weiss, G. H.: Spatial distributions 
of homes for journeys to work by different modes of transport. Transportation 
Research (to appear) . 



46 



July 1, 1973 throuei^ June 30, 1971+ 

PLuLIC litALTH i;ERVICE - NATIOriAL lUSTIlUTES CF HEALTH 
Dl Viol UN OF COMPUTER REbEAFXfl ANP TECMi-JOLOnY 

1. DCRT b 

Serial iJumbrr 

2. hLURISTICo LALOR/vTORY 5. James H. Slagle 

Chief 



Heuristics Laboratory hiad a very fruitful yi^ar. More 
Lfian twenty paper:s arc either puoli.shed rr accepted, aocut 
three papers per non-year. Most prcblens undertaken by the 
stcfl ot the Laboratcry are concerned with the application 
or computers to oioi.iedical fields. Since these problems are 
lar^t: and dirricult, efficient anri heuristic rnethoJs are 
needcc to solve theiii. Therefore, the activities of the 
Laboratory are creative anci at the rrontier of computer 
application, and are different fron those routinely 
inpl er^ent i nt existing r.ethcds and systeios. The projects 
worked upon uy the Heuristics Laboratory during the fiscal 
year are sumnarlsed as follows: (The names of nembers of 
tlie Laucratory involved in each project are listed in 
pa rerithes i s . ) 

A dti; LESIGi'J FCK CliEf'JLAL STPUCTCf^t SFA^C!' (I'cdes): 

V('e are collaborating with tlie Walter Reed Army 
instituLe of Research (wIvAin) on the desi&n of a ctien-ical 
substructure system in conjunction with the conversion of 
the obsolete t.l'.AlK systeni froi'i a second reneraticn t/^ a 
tni rd ..erieration coi'iputfr. The nev' --yster^ v; i 1 1 allo'>' 
searching oi-line for both structiire and sl'os t ructut-e, and 
vill =!CconTnodati^ 3 file '.'hose co'^rif unds can number in tt-.e 
.nilljcns. There are tv.-o oasic nev.' fnements in cur desi:in. 
First is the establ i shtnen t of a sy^te;^ (-f structure fragment 
scree^.ns. t^e have cevelrped a liniform, iterative methor' for 
deteri'inins a complete set of screens froti sin:-^lc atones t^ 
quite lar,,e Iragrnents. Second, i ;; a new use ■■f ha^h crdin^^, 
v/hic(' ror trie first t\..)e brink's to bear tfic full speei.'. of 
ranco,',) access on suus t rue turt: searchini-. These two elements 
have ueen i nte j^ra ted into a rational f rai^'C v,'ork based on new 
theoretical work on the discrimination nov.er of screens. Ive 
riave ootained an ele^^ant and practical design vdiose 
performance is versatile, precictable and e'f^iicient, and an 
improvement over other chemical search syste^.s. Preliminary 
results on estaul i sh i ri^ screenin'^ fra>/ments iiave been 
grat i f y i ns . 



47 



JPtcrWAL cjLARCM (-ieller) 



Ab of ijeprenber i, i'j73 the -IIM Mass Sdcc Searcii System 
(;;So-i) wao officially nut on the international "lE cor-puter 
nav.iork by the Unitei Kin^ion Spec Data Centre. M I M ' s rcW^ 
in the future will be that of a consultant for the syste--. 
A copy of the systen- regains nt ."J I H for internal w^c an-i 
further research and devel opnien t . 



a no 



Hel ler) 



TATICm of Kao:. ^P-£C"i PiJi'ET'iY OATA (Slagle, Chan^ 



to reors^anize a larp-.e ■'^ i 1 e of 
the vTiis'-lay of the reorn;an i zee 

ne\. i!;rv Lips '~v cO'-ipounus iiy the 

of certain nr:iss sof^ctral peaKs. 

for four rroups:- thiolesters, 
sulfoxi-Jes, aromatic caroamates, and a.^ino esters. The 
definition rules for these nrouos have been testevi 
intensively, and found to be aule to d i sc r i n i na te a c-'ipounH 
in tlie 'groups from other coripoun-is not in the rrouos. 



A prof.rain vvas developed 
mass spectral .^ata. Fro.T 
data, we -were aole to define 
presences and the absences 
oo far, ive have good results 



f.:LAD - AiJ Ud-LINE i-:ODrLinG LABORATORY (Knott): • 

i.LAB lias been further develot>ed. We have spent 
considerable time promoting MLAb and crl l ab(;rat i n^^ v;ith 
users in setting up models and applyinp; then. The third 
edition of the MLAi3 manual has oeen prepared. Uslni; MLAG 
jointly with Jean-iiarie Ketelslager, v/e have obtaine-! sone 
results on hor.Mone oindin^. 

[lICROblOLOGY UmTA bAiiK (f'orton): 

Collaborating vlch Dr. Krichevsky of iJID.^, we revised 
the prG;;rani for inputtinfi Jata to the i li c rob i ol o^y data bank 
to ;,ive consideraoly nore extensive diapnostic nessas:es in 
cases of deviations from the snecified input fornat '.jhich it 
will not forg i ve. 

TloSUL Trpl.'iu COMPUTER a\Do (Tixon and Norton): 

we are col 1 abora t i nj:, with Or. Kavhoe of rJI A I D on the 
evaluation of the 1330 trays usin.^ the t i ssue- typ i n? 
pro«rai:i. The results dencnst rated new pattern<^ 07 
serun-ant ir,en specificity on the ij I H tray. That Is the 
prcAramenacled many "false pointers" previously interpreted 
as noise or laboratory error to be ' ^nter.^reted 
systematicully in terms of (scmetiries i.eak) serum reoctions 
to anti.\ens determinants. 



48 



COMPUTER MODEL OF THE HUMAN MEi.'STRUAL CYLLE (Dixon): 

vje are collaborating v,/ith Dr. Car^-illc of NICHD on the 
modeling of the human menstrual cycle. A con-'puter m-.-jel 
written in the DYNAiiO 1anrua>?,e simulates the pr^icess of 
ovalation and the variation of hormones ourintr the human 
menstrual cycle. Last year the model J Id a -:ood jou f)f 
slmulatlni^: the normal cycle and a dynamic mechanism for 
atresia vjos postulated and put into the model. This year 
the model was completely re'.vrltten. Many av/fn.ar-; equations 
were streamlined and the whole mcJel v.as made simpler and 
easier to understand. This nev. version of tho model v/as 
called flAl^.K't. A simulation of pregnancy '.;as also includei 
in MAia<'+. >-. statistical version of the -.;cdel called .''APK5 
v/as also written. MARKS can be run for many cycles and 
statistics of hornione levels at various times durinp: the 
cycle can be collected automatically. MARKb was also 
written. This is a version of MARKS v/hich uses a larger 
time Interval (DT) so long runs can be made at less expense. 
A set cf about 10 validity tests was then prepared. The 
model passed half the validity tests vntiiout chang:e. 
Modifications were then made to the model to .;et it to pass 
the rest of the tests. This v;orr< is still In progress. 

COI'^PUTER INFERENCE AND LEARIvING (Dixon): 

The soal of this project Is to develop a program to 
aiscover relationships in numerical data. It Is hope i tlsat 
this program will be of practical use to research workers in 
all fields for automatic analysis of experimental data. The 
basic idea is to improve the standard technique '-t linear 
multiple re;';ressIon analysis .jy usin--, products of predictors 
as if they vjere nev; predictors. One mi_;ht call tiiis 
technique non-linear res^ression analysis. Heuristics arc 
necessary to reduce the very larr:e number i. f 0033 i >le 
combinations to a reasonable number. A basic orog-am called 
;aC1D2 '.jos v.rltten. ACir.2 d<:>es a multiple re-^ression 
analysis on a given data set by the steowise method. All 
products of preiMctors are used. The '-'c'els thus ob>taIned 
are then tested on a portion of the lata nc^t used In the 
regressifn analysis. This Is called ;.ne acid te-^t - hence 
the name or the proiiram, A n'^muer of auxilliary nrr^rans 
have Deen written to collect, faorlcate and manipulate data 
sets vrr this pror,rar'. There ere no\ iJ dat? "^.ets on which 
to test this prosra(:i. They Incli.'de bloorl oressure data, 
tissue typlne data, stock market averages, a corv^jter 
simulated vehicle, a computer sli^ujlatPd radio, tables nf 
random numbers, as well as others. The AC IP? program is in 
a pri.iiitlve condition. No heuristics are us*:" J at nres-^^nt. 
It is hoped that the model v.Ill oe able to learn its own 
heuristics. But preliminary results indicate that 
non-linear regression works better than linear regression. 

49 



auTOi.aTIC PROGtUi-'i-VcRI F ICATION (olasle, Chanf: and Lee): 

The application of theoren provin-^, tpchniques to 
prograr;^ ver i f i cat i on was cons i Je re i . A new an-! direct 
technique for proving pronrams correct was <-btaineu. 
Prccress in this field way eventually enable us to test 
wfietner softuart is reliable or n^t. 

Aui'Ji-iATIC P-^.VTER'' RECOaiHTIOii (Slagle, Ghana; an. I Lee): 

i'.n algorithii for finclin"; prr-totypes for a nearest 
neiL;noor classifier v;as cevelope.i. The a1;oriLhr,i ;vas 
dpplied tc bik cases ot liver uisease, an j only 5k 
protot>pes v^ere found iiecesi^ary tc achieve luU .^ of correct 
a i agriOc> i b . 

In ado it ion, nany clustering.; a r.,o r i tiir is v/ere tested. 
The short spannint^ path al'^orith"! v.-as found tr be si:.-.pler 
ana co.'.ipa raol e w'ith the ,.~ii n i f'Ur.i spanning tree iiethod. 

PROFESoIGNaL ACTIVITIES; 

J. R. Slajile participated in tf.e National Institute 
of Education Plannin.fc Conference on Produc i t i v i ty and 
Efriciency in Education in the united States. Me frave a 
computer science ser.'inar at IB'' Research Center at Ynrktown 
Heights, tne State University of Nev York at Albany, and the 
University of Texas at Austin. 

C. L. Chan.i ^ave sen.inars at the Ruf^.ers University, 
IdH r..ese3rch Center ?t Yorktov.n Heights, an^i IBM Research 
Laocratory at San Jrse. He tau,^ht vMtli Slai-le a nCP.T course 
on pattern reco.^niton. 

J. K. Uixon taught tv.": DCPT course^ on heuristi: 

proc,rai lii.i n,-. 

^' ''^« tieller .iave a lecture at the 'JATO AS I on 

CGr;pucer .representation and ^'an i p 1 1 a t i on of Chemical 

Inforcation" in rlolland, and a talk in the CODATA meetin,-^; in 
Freiourg, Gerna.Ty. 



L. Hodes j:ave a seminar at the University r^f Maryland. 



50 



b. D. Kiioct taught hLhS and the -.raphics courses in 
DCt;T. !:e has written o r.onchs i;orth .^f liJTEHFAC^ articles 

en varicus topics. Also, he tautht tlie computer science 
component ':n the Computers in Clinical r;edicine orogra-n 
sponsoreil oy DLRT. 

P. C. T. Lee i:)resented a paper at the International 
Syniposiun on Coriputer and Chinese Input/Output Systeins at 
Taipei, Taii^an/ ana a paper at the Very Hio:h Level Lan^uai^es 
oyii-iposiun at Santa Monica, Calif. Me ^ave sen'inars 'at the 
Temple University, Philadelphia, the Universities of Texas 
at Austin, Dallas and Arl i n?^ ton, the University of riichii;an, 
and JorLh Larolina State University. He tau;ht a DCRT 
course on cluster anal y^^ is. 



L. jorton participated in a 
on Ditj syster-;s, givina; a 1/ z 

system. 



^ove rnmen t-w i de b r i e f I ng 
ho u r talk "n his i n ite x i n p; 



PuuLICaTIUNo: 



lu 
li 



Chan 

Disc 

Ho. 

Chan 

Clas 

llel 1 

Spec 

Hel 1 

of f". 

Anal 



t;/ c 
r i m i 
'J, P 
f,/ C 
sif i 
er, 
t ra . 
er, 
ass 
vt ic 



. L 
nan 
p. 
. L 
ers 
i> . 
" T 
s . 
Spe 
al 



as 



"Pa 
Func 
9-6b 

"Fi 
To 
, "C 
appe 
/ C. 
rome 
emi 3 



tte 
tic 
Z, 
nd i 

ap 
omp 
ar 

L. 
try 
t rv 



:!el ler. 



1 no 
Hell 
of ,\ 
Meli 
"A L 
ot i 
Houfc; 

[•;aa i 
pp . 
iCnwt 

d.-.C^ 
!;nc t 
Comp 



tlvol 
e r , 
cjrep 
er, 
onve 
per. c 
s, L 
a c I o 
Ibl- 

L, L, 

/ Vo 
t, G 
uter 



ut I 
S . 
hri 
s. 
rsa 
ra 
• / 

n T 
iiJb 
. U 
1 . 
. L' 
Jo 



17 



/ R. 
e, " 

of 

/ I- 
and 
, D. 
ona 1 
d D i 
em i a 
at me 
Jan . 
"A 

, ;jo 

"Ha 
nal . 



J 

A C 

a S 

Ka 

Do 

Kg 

fIS 

ss i 

uto 

nt 

Id 
du:'; 
. 1 
Shi 



Recognition by Plecewise Linear 
, " IEEE Trans. Computer , Vol. C-2 2, 



Pt. I^i7 

Protot 

a r i n X 

er Tech 

h\J() A 



PCS i''^! 

£F. Trar 



Ne 



1 ques 
I Proc 



for 
eed i 



a rest !Je i shbor 
on Cor-puters . 
I nte rpre t i n;r Mas 



h a n >j; , a 
ata UsI 
rlay, 13 
el dmann 
versat I 
ten," J. 
, and A 
mi ne , " 
ver, H. 
earch S 
1 a r i t y 
tic Opt 
ann i n>:. 



d K. C 
r. C 1 Li s 
k. 

H. ;■;. 

nal f'S 

Chen. 



*ol 



Ja 
Ph 



hu, 
tcr 

Fal 

Sea 

Poc 

cobs 

arn. 



"The Interpretat 
Ana lysis," 



fi. Fal 
stem, 
omoa r ! 
m i za 1 1 
Radio 



es, 
III. 

son . 
on o 
1 o p. y 



es, 

rch 

r.n, 

V d 
and 

Di 

" I 

f E 
. V 



and 

Syster , IV. 
13, 130 (1973 

"A m study 
, Ui>D-i+dC, (1 

G. W. A Mi In 
splay and Pic 
n press, 
xternal Eeam 
ol . ll'J, !Jo. 



rin-.; Systci^ lor Cor.ib i na t ions, " 
i-ip. '*:)-'j'^, Jan. l.r/'i. 
Functions." To appear in the Cr i 



s 
i on 

). 

973) 
tt in, 

1, 



Krichevsky ana L. Jorton, "dtora!i,t: anu Manipulation of 
L'ata uy computers for Determinative i'ac ter iol ocy . " 
To appear in Int. J. of Syster^atic Sacter i ol o;;v . 



51 



12. 
13. 

Ik. 

lo. 

17. 

13. 

1j. 
2U. 

21. 



Kricfievsky/ and 
Interacting vv 1 1 
Int. Congress t" 



Krichevsky, and 
2S an I n1 ormat i 
ori Cul ture Ccl 1 



Lee, R. C. T., 
iipannini Trees 
SvnpC'S i Uii'i on _C_o 
Au>i. li;7 3. 
Lee, f^ C, T 
Improved Projsra 
CaCM , Vol . 17 , 
Slafile, J. R,, 
o impl i t" I e rs / Co 
In JACf'i . 
Si a^le, . R./ 
lincvcl oped ia of 



L. Norton, "An On- Line i 
1) Bacter iolo;-. ical Data," 
or 8acter iol CSV . 

L. Norton, "The l.'orld's 
on Systerr,," Pr<jc. r-i Znd 
ect ions . 
and C. L 



'uery P roe ran fcr 
Proc. o"" First 

Cultijre Collections 
I nt. Conference 



Chans, "Appl ica 
to Inforr-at Ion Storage," 

liiputers and Chinese Input 



t i on or f i I n i r,urn 
Proc. of Int. 
/Output Systens , 



Si agl e, J . R., 
v/ith Some Clust 
Pattern Reco^n i 
Slagle, J. R., 
To appear in En. 
Sla^le, J . R., 
Theorem Prover 
CACli , Id, pp. b 
Slagle, J. R., 
for the Theorie 
appear in Compu 



C. L. Chanty, and R. J. V'a 
m-Synthes i zin-^ Al gcr i tnm 
:Jo. '■4, pp. 211-217, April 
"Automatic Theorem Provin 
Mi.iutat i V i ty , and Associat 

"Theorem Proving, Autonat 

Computer Science . 
C. L. Chang, and R. C. T. 
er Analysis Algorithms." 
t ion . 

J. Dixon, and T. Jones, " 
cyclopedia of Computer Sc 



Idln-er, "An 

and its Correctness," 

19 7U.__ 
^. for iheories vnth 
ivity." To anpear 

ed." To appear in 

Lee, "Experiments 
To appear in 

List Processing." 
ience . 



and L. Norton, Experiments with an automatic 

having Partial Orderin.: Inference Rules," 

82-6^J8, Nov. 1973. 

and L. Norton, "Autom.ated 

s of Partial and Total Or 

ter Journal . 



Theore-i-Prov i ni 
derina," To 



52 



July 1^ 1973 through June 30, inyu 

P(IS-NIH 

Division of Computer Research and Technology 

Summary of Branch Activities PCRT 

Data Management Branch J. Fmmett V7ard 

Branch Chief 

i . SUMMARY 

New computer software capabilities had a significant impact 
on systems developed by the Data Management Branch during fiscal 
year 197it. Although the new ARMS system v;as not the most 
complex system developed in the branch during the fiscal year, 
it certainly was the one which best demonstrates the advances 
made over the past year. Virtually every personnel office at 
the NIH is entering its data at a communications terminal and 
each of these offices can retrieve information at the same 
terminal only minutes after a request is entered. 

This same approach, using much more complex programming 
techniques, has been used in developing a retrospective re- 
trieval system for clinical pathology laboratory data. Requests 
In prior years were held until there was enough volume to 
justify an expensive computer run. This often meant several 
weeks of delay In obtaining valuable research information. Now 
It Is possible for a research investigator to enter a request 
one day and receive the appropriate laboratory data the next. 
The most significant aspect of the new approach Is that the 
cost of obtaining this Information has dropped from approxi- 
mately $500.00 to $20-$30 per request. During the next fiscal 
year It is expected that this same technique can be used to 
Interface other clinical data with the laboratory data to render 
It more useful for analysis by Investigators. It should also be 
pointed out that this methodology facilitates data security for 
the Office of Clinical and Management systems in the Cllnlcai 
Center by enabling terminal control over laboratory data 
users . 

In the highly dynamic wcfld of software support, the branch 
has embarked on an effort which will combine each of Its 
generative systems into a completely integrated data management 
system. These systems are extremely useful as individual 
entitles/ but require several refinements ard much more inte- 
gration to keep pace with the latest facilities avallBble a'^ 
the DCRT. As part of this effort, the Recursive flacro Actuated 
Genarator (RMAG) Is also being completely modified to improve 
Its Input and output capabilities. 



53 



The Biological Abstracts biweekly HIOSIS service was sub- 
scribed to by the DfIB this year as a supplement to the existing 
Chenical Biological Activities Service. Both services are pro- 
vided free of charge to the NIH community. During this fiscal 
year vje have also negotiated with the NIH library to advertise 
both services and to provide assistance to prospective users in 
preparing profiles of search requests. This relationship seems 
to be working out quite well. 

The Harwell mathematical subroutine library was added to the 
DCRT Mathematical and Statistical Program Library during the 
year. These subroutines represent several years of ' programmi ng 
effort by personnel at England's Atomic Energy Commission and they 
should add significantly to the flexibility of the DCRT library. 
A joint effort has also begun with library users to determine 
what new facilities should be made available and what old pro- 
grams and subroutines should be eliminated from the existing 
library. it is planned that better communications terminal 
facilities will be added during the next fiscal year. 

Under the direction of firs. Prev/i tt and with the continued 
support of the National Bladder Cancer Project, the PEEP/nECIDE/ 
CRAPH general purpose conversational programming system for syn- 
tactic and semantic Information processing on structured data 
bases has matured materially and lead to positive research 
results. This general purpose system is being used to solve the 
fundamental problem of classifying patients for treatment and 
tumor type; localization, extent and progression constitute 
indl spens i bl e information for indicating the modality of choice 
and the scheduling of therapy. Our role is to develop a hither- 
to unattempted objective and quantiatlve basis for bladder 
hi stopathology, which is currently regarded as the "penultimate 
truth", clinical Information, cystoscopy, and exfoliative 
cytology notwithstanding. The data bases consist of stained 
clinical biopsy specimens, ^canned and digitized at high photo- 
metric and spatial resolution and magnification at JPL v/!t^ 
several visible wavelengths and white light, and analyzed at 
DCRT with the PEEP/DEC I DE/HRAPH systems. 

Ctrlng the past fiscal year, PEEP has acquired new built-in 
picture operators for global dynamic thresholding based on 
several prlncioles; spatial derivatives; Laplacians; fast and 
ultra-fast Fourier and Hadamard tranforms and inverses; sub- 
picture creation automatically, by explicit sampling and Inter- 
actively via displays; masking; convolution; connectivity 
determination; medial axis construction, and shape, size, 
brightness, integrated density, and orientation descriptions. 
'./ith these capabilities we have been able to perform picture 
segmentation optimally In the presence of a-tifacts and Inter- 
cellular debris, to extract cell nuclei, describe nuclear 
features quantitatively, and assemble attribute lists for 
pattern recognition and multi-dimensional scaling purposes. We 



54 



plan to process 100-220 cells automatically by June, covering 
the range of Grades \, I! and III papillomas and papi 1 larycar- 
cinomas. DECIDE facilitates decision-making and pattern recog- 
nition by addition for facilities for defining sub-sets, com- 
plexes and syndromes, for statistical decision theory (Bayesian, 
maximum likelihood, other; linear and quadratic), and for 
tracking error rates. GRAPH, a new system component, provides 
soft- and hard-copy two and three dimensional displays of curves, 
graphs and surfaces. Commands are rel at i vi s t i c; that Is, they 
can be issued from the vantage of the viewer or the viewed. 
Several publications and public addresses on the system and its 
results for bladder cancer h i stopathol ogy have been issued. 

During the coming fiscal year, we plan to unify the three 
subsystems, PEEP/DEC I DE/HRAPH, into one system on the PDP-10 
dual processor, 'ie plan to process and scale or classify cells 
and tissues in several dozen scans (amounting to several 
thousand cells). We v;i 1 1 attempt to optimize staining by 
further experimentation with Harris hematoxylin, Feulgen, and 
ga llocyan I n-chromal um, and to optimize mul ti spectral scanning 
wavelengths for resolution of contrast and characterizing 
morphological detail. 

In a highly developmental effort the Data 'lanagement Branch 
Is collaborating with Dr. Pinesh Sharma, NHLI, to facilitate the 
planning and forecasting mechanisms within the OD, NHLI. The 
basic objectives of this system are to support both strategic 
and operational planning, to assist In the budget staging and 
monitoring process, to develop future projections based on past 
trends and to provide a better method for evaluating manpower 
and other resource allocations. 

The methodology for establishing such a facility will be to 
combine much of the existing MIH grants, financial and budgeting 
data with information endemic to the NHLI and to provide this 
information in summary fashion to the Office of the Dit ctor. 
Facilities such as a flexible terminal based interactive system 
will be made available to enable quick modification and retrieval 
of data in formats dictated by the user at the time of access. 
i/e have high hopes that, through the efforts of Dr. Sharma and 
the programming expertise of DMB personnel, the sys'.em will be 
highly usefjl to the NHLI. 

The following Is a summary of those systems which 
were developed by the various sections In the branch and, 
where appropriate, an abstract of those projects which 
demand a more detailed explanation. These projects can 
be summarized best by organization structure. 



55 



Documentation and Systems Support Section 

1. Clinical Center Data Processing Support 

Provided reports to Clinical Center investigators from the 
Laboratory Data Files (BETA) at their request and according to 
their specifications. 

2. Case 1973 Reports 

These annual fiscal year reports vn 1 1 contain 135 different 
listings, tabulations and ranking tables to be published in 
five volumes by the OD/ADPPF/ORA. For Fiscal Year 1972 we began 
to standardize the approach to producing output tables. lie 
achieved only partial success In that 1/3 of the 120 case reports 
were produced without programming. By using these program gener- 
tion techniques It Is expected that we will avoid reprogrammi ng In 
several new areas this year. Eventually It is hoped that program 
generation will significantly reduce the time and resources 
necessary to support this annual requirement. 

3. Trends In Graduate Enrollment and Ph.D. Output In Selected 
Science and Health Professional Fields. (1962-1963 thru 
19701971). 

k. Redesign of ARMS Personnel System 

All institutes without appointing authority are currently 
using a new terminal based personnel data collection and re- 
porting system. The new system is being integrated with the 
departmental Terminal Data Collection System and it is antici- 
pated that all of the personnel offices at tJ I H will have the 
full power of the new ARMS facility by the end of FY 197ii. 

5. Vitamin C Study, C.C. 

Acdltional relationship tables were provided at the request 
of Dr. Chalmers, the former Clinical Center Director. 

5, Clinical Center Census Reports 

Institute, Ward and Bed complements were changed effeotiv? 
lO/ni/73 and the system was modified to reflect these changes. 
An additional program has been written to provide a recurring 
monthly Geograpnic report. 

7. Diagnostic Radiology Department Case File, C.C. 

This computer system was developed to replace a manual 
radiology record keeping system. Neither the old manual nor 
the new automated system records diagnostic information, but 
provision for doing so is possible. 



56 



8. Physician Authority List System, C.C. 

Additional programs were v/ritten during the year to provide 
Index cards of physicians involved in patient care. These 
programs are capable of terminal communication. 

9. Fogarty International Center 

Data on Russian Scientists, Doctors and Institutions 
was collected by Library of Congress personnel using prompt 
programs written by DMB. A "Directory of Institutions" and a 
"Directory of Personnel" has been produced from this data. 

10. NHLI Information System Phase I! 

System established for collecting, editing and updating 
of all MHLI extramural programs using data from both the 
Division of Research Grants (DRH) and the NHLI. 

11. Study of Baltimore Cancer Research Center (BCRC) Time 
Sharing Requirements 

At the request of BCRC, the Data Management Branch conducted 
a study of the possible alternatives to BCRC's present use of 
Tymshare's computer system. A GSA requirement precludes BCRC's 
use of the Tymshare service after October, 1P7U. Results of the 
study indicate that some use of the Tymshare computer system is 
necessary until a table making facility can be duplicated either 
in-house or by another time sharing service. It Is anticipated 
however that Cai 1 -a-Computer v/i 1 1 provide much of the new time- 
sharing support for BCRC and that two of the larger systems, 
namely Census and Microbiology, will be supported on the DCRT 
computer system. Work is In process now to accomplish this. 

12. rjC I Contractor Data System 

An automated MCI contractor data base was created using data 
from the DRG Inpac system suppl e.nented by additional elements 
entered by the NCI. The Research Contracts Branch now has o 
facility for automatic reporting of both open and pending 
contract Information. 

13. NCI/DMB Contract Involvement 

Personnel of the section are acting as consultants to 
various MCI personnel and contractors. These Include contracts 
with Tracor/JI tec, JRB Associates and TRW. 

m. Opportunity Skills System, ODA, P 



57 



15v Results of Questionnaire on Housing, Transportation and 
Child Care Needs of NIH Personnel, ODA, P 

16. Case 1973 Data Preparation for NSF, OD/ORA 

17. NICHD Grants System 

18. Blood Assay Test Results, NIMH 

19. Lupus Data, NIAfIDD 

20. Computer-assisted Electron Microscopy, NHL I 

21. KWIC Indexes for the NIH Central Library, DRS/L 



58 



Applied Systems Programming Section 

1. Type II Intervention Study. 

The system supplies all of the data storage, validation, 
purification, monitoring, and reporting capabilities required 
in support of the Type II Inverventlon Study being conducted 
by the Lipid Metabolism Branch, NHLI. The reporting capabilities 
were added this year and the OMB is maintaining the present 
system. Analysis of the data is partially defined and will 
continue thru FY 75. 

This system makes it possible for Lipid Metabolism Branch 
personnel to store and to subsequently do reporting and analyses 
on their patients with blood lipid disorders and to relate these 
analyses to other family members. It also provides them with 
listings used as an aid to the doctors for patient handling in 
the clinic. Of particular current interest is the development 
of reporting capabilities to serve the Patient Safety Monitoring 
Committee. Support for the system and training of flHL I personnel 
in its use is supplied by the DMB. 

2. Carcinogenesis Bioassay Data System 

This computerized data processing system makes it possible 
for the Office of the Associate Scientific Director for Carcino- 
genesis, MCI to control the data acquisition, input, purifica- 
tion, reporting and analysis of animal research data related 
to the study and Identification of various agents In order to 
determine their carcinogenic effect or capacity. The Study 
includes many thousands of animals being tested under various 
contracts awarded by the NCI. Examples of output are a summary 
report, pathology report, survival curve graph, and weight curve 
graph. Output can be produced in hard copy or on Micro-flche 
for all, selected contracts, or selected experiments in a 
study. The contractors can, upon request, be provided with a 
complete data base subset of their data. In addition, the 
Branch supplies assistance to contractors before and after 
their studies are included in the system. 

The major data management system developmental effort has 
subsided, however, the DMB continues its support. The 
attention has now shifted to the analysis phase and the DMB 
Is actively involved in this. 



59 




automatic SNOP encoding of clinical and pathological diagnosis, 
error correction, edit/update, a fixed format report, and a 
compl ete 
which pi 
quer i es 



;orrection, edit/update, a fixed format report, and a 
:ely generated COBOL retrieval program executed via TSO, 
)rovides almost complete flexibility in composing quick 



k. Computerized Distribution List. 

This system will provide the Grant and Contract Ouide 
Distribution Center, DRG, with the capability to create, main- 
tain, and selectively produce labels to be used in the distri- 
bution of the Grant and Contract Guide and/or any of the 
various supplements. The data base will eventually be augmented 
to Include other NIH mailing list requirements. 

5. NIH Appropriations and Obligations Information System. 

This computerized data processing system provides the Office 
of Legislative Analysis, CD, NIH, with the facility to maintain 
a data base of NIH Appropriations and Obligations by total NIH, 
total NiH minus programs transferred out, by activity within NIH 
and by activity within program. 

6. HIH Personnel Accident Reporting System. 

This system utilizes the computer to monitor and retain a 
complete history ov NIH employee accidents, fol low-up-ac ' ions, 
and corrective measures taken. It communicates with the DHEW 
computer system via magnetic tape. 

7. NIH Space Management System. 

The Space Management System helps the Space Management 
Office manage NIH space utilization and allocation. It also 
assists other offices by maintaining Information relating to 
areas containing biological or radiation hazards. 



60 



S. Iledlcal Records Discharge Diagnosis System. 

The system provides the Clinical Center Medical Records 
Department with the ability to maintain a current/ accurate data 
file of all Clinical Center in-patients' discharge diagnoses. 
It also supplies various useful reports and listings In the 
form of indexed MICROFICHE. 

9. NIH International Activities and Personnel Monitoring System 
for the Fogarty International Center. 

10. Type V Intervention Study. 

This system provides Or. Hreenberg of the Lipid Metabolism 
Branch, NHLI v;ith the capability of storing and purifying for 
analysis data collected by MHLI on patients with TYPE V lipid 
disorders and their first degree relatives. 

11. NIH Central Registry of Biological Agents and Materials. 

This system Is designed to aid the Environmental Services 
Branch, DRS, in its administration of the Registry. The primary 
purpose of the Registry is to make feasible the monitoring of 
hazardous and potentially hazardous materials and agents In use 
at IJIH. 

12. Pharmacy Computer File System. 

The Clinical Center Pharmacy maintains a complete, accurate 
file of relevant information on all drug products which It has 
available. From this file we produce a variety of reports of 
use to the Pharmacy Itself and to doctors and nurses who pre- 
scribe and use Pharmacy drug products. At the same time, the 
file serves as a centralized, computer readable file of drug 
Information about drugs used in the Clinical Center. 

13. Dental Clln«c Report. 

This is a very basic system that aids the Clinical Investi- 
gations and Research Services Branch, NIDR, in maintaining a 
computerized cata base which reflects all dental services 
rendered by this Branch for ell Institutes. 

It*. Fgytlan Autopsy Data Analysis. 

Dr. Cheever of the Laboratory of Parasitic Diseases, NIAID, 
performed approximately 250 autopsies on subjects In t^gypt. He 
now wishes to computerize the data in order that an analysis can 
be made relating to a specific parasite/parasite egg resident in 
various organs of the subject's bodies. 



61 



15. ML-A Data Maintenance System. 

The nature of Dr. Rogentine's study for the Immunology 
Branch of NCI Is to try to determine if there Is any connection 
between specific HL-A types and disease. Now that the system 
Is operational, our role is supportive. 

16. DCRT Personnel and Training System. 

17. Leukocyte Infusion Update and Reporting System. 

The system collects data and reports on Leukocyte Infusion 
information for the Medicine Branch's Leukemia Service In NCI. 
The information collected Includes specific infusion Information, 
laboratory follow-up information, and infusion series summary 
Information. Data is extracted from two related systems, the 
Cell Collection and the Donor Recipient Information System. 
These data are then combined with the Leukocyte Information 
for other reports. 

18. Lipid Protein Study - Lipid Metabolism Branch, NHL I 

19. Hypertension Study Data Processing System. 

The purpose of the study is to try to find correlations 
betv/een hypertension and other variables such as the kidney's 
secretion of different levels of Renin. This effort supports 
the Reproduction Research Branch, NICHD. 

20. CI In/Path Test Result Extraction for the Immunology Branch, 

NCI 

21. fiVIF Medical Records Information System. 

The Emergency Virus Isolation Facility, HC I , has requestad 
that its data collection and reporting system be made con 
versational . The effort essentially is a redesign of the existing 
system to provide total data control within the cor Fines of the 
Building ul facility. It Is hoped that when *:he sy«. t'^n is com- 
pletely operational, the loss risk relating to sensitive employee 
data will virtually be eliminated. 

22. Animals for Research. 

At the request of the NCI, DMB is developing a completely 
automated system for the Institute of Laboratory Animal 
Resources, (ILAR). The Intent of the system Is to Improve the 
publishing and referencing facilities available from ILAR, which 
publishes data relating to sources of animal stocks, fluids, 
tissues end organs. 



i 



23. Cigarette Condensate Study System. 

The study being conducted by the Ftiolgy Branch, MCI and 
executed by Hazel ton Laboratories is an attempt to determine the 
possible alternatives for people who must smoke and to establish 
patfiways of minimum risk. Df^B provided programming support for 
the data collection and reporting segment of the project. 
Analysis will be handled by riC I personnel. 

2k. Computerized Grant Supported Literature index. 

At the request of the Program Analysis and Evaluation Branch, 
MCI, the DMB is providing a computerized capability to 1) create 
and maintain, 2) index on keyword, and 3) print for publishing 
the Grant Supported Literature Index. 

25. Computerized Patient Menu Labels. 

This system v/ill search the Admission file for active In- 
patients and produce three labels for each. The labels will 
be placed on each of the 3 segments of the patient's dally 
menu for identification of the meal selection by patient. This 
will save the Nutrition Department in the Clinical Center from 
having to hand prepare approximately 90n labels per day. 



63 



Scientific Application Section. 

1. Psychophysiological Measures 

In support of the Laboratory of Socio and Environmental 
Studies, NIMH, the Data Management Branch has developed a 
generalized data management and statistical analysis system for 
psychological and physiological measurers. During this fiscal 
year, an artifact detection and removal module for heart rate 
data was developed, the system was completely documented and 
the user/system interface using live data was started. 

Future efforts will concentrate on the integration of 
laboratory processing data with the computer and on the possible 

development of a file maintenance capability. 

2. Data Base for Surgery Branch - — 

The Surgery Branch of the NCI had a need for a data 
collection and reporting system. This included the 
editing, storing and retrieving of all data related to re- 
search projects of the Surgery Branch, NCI. 

During the past year two new protocols were Incorporated 
Into the system. To date, the DMP has provided CPS data collec- 
tion/edit prompt programs as well as an update, retrieval and 
reporting system for nine NCI protocols. Data files have been 
developed to include one file of common data and an associated 
auxiliary file. This represents a file of all Surgery Branch 
Operations Including those which have not been expanded Into 
separate protocols. Nine additional protocols for which only 
core data will be entered are currently being Implemented. 
Analysis of the data may lead to more complete protocols for 
operations not under special Investigation at present. 

3. Car.cer Survival System 

NCI 

The Survival System was orlglnaTiy developed to 
support the End Results In Cancer Studies. Maintenance 

and Improvement of the system is the primary goal. 

During the year several more requests were made for copies 
of the system. In addition the.'e has been Increased use 
of the system by elements outside the Cancer End Results 
Group. A number of minor problems have been solved. 

A new submission of cancer data took place this yeaf and 
the CER personnel ran their own edit and analysis programs 
w th some consultation with us. The next submission of data 
will Involve substantial changes in format, making It necessary 
to make changes to the Survival System to reflect format and 
coding changes, 

64 



h. Current Awareness 
(CBAC) 



Search of Clinical Biological Activities 



The biweekly issue of the CBAC data is received from 

the Chemical Abstracts Service (CAS) In Columbus, Ohio. 

These issues are used to supply the NIH community with 

information on current articles related to the chemical- 
biological area. 

Retrospective searches are also run monthly or on demand 
In special cases. 

The major non-routine effort in this area during the year 
was conversion of all PL/1 programs in the system to the new 
PL/1 compiler. At the same time, the print program was 
streamlined in order to realize additional savings. 

5. Current Awareness Search of Biological Abstracts (BlOSiS) 

This is a new service, offered for the first time at NIH 
in 197U. Tv;ice a month tapes will be received from the Biological 
Abstracts Service and Information will be disseminated to the 
NIH Community thru the same vehicle as CBAC. 

The major effort went into the development, programming, and 
testing of the reformat program for this new data base. The 
original CBAC search program is used for BIOSIS data after 
reformat. 

6. Aortic Valvular Disease 

In collaboration with Dr. Walter Henry, NHL I , the DMB 
is providing programming support for a study which will 
attempt to define prospective pre-operat i ve risk for heart 
valve replacement patients. 

dub's main involvement with this project during t!ie past 
year has bee-i to respond to questions concerning the use of 
the data collection and file maintenance programs we provided. 



65 



7. Cutting Oil Study 

This system supports the efforts of the National 
Institute of Occupational Safety and Health. It is a study 
relating job type to mortality and morbidity of cutting oil 
workers. A data base has been established, and a survival 
analysis program similiar to that developed for the Steel- 
workers Project will be written. 

During the latter half of 1973 a number of frequency tables 
were produced as a preliminary aid to the development of speci- 
fications for the analysis. 

8. Federal Women's Program Statistics 

The Department of HEW supplies the Federal Women's program 
with tables reflecting NIH employment and promotions by race, 
sex^ and employment level. Data is extracted from these tables and 
used to create some graphs. A Calcomp program was set up which 
allows the production of several different graphs and bar charts 
as desired. Complete documentation for running the programs 
was given to Ms. Adele Nusbaum, 0, FED. 

9. Gastroenterology Training Review 

A coding technique for recording the training, education, 
and experience information on questionnaires sent out to training 
grants recipients was devised. After the data v;as keytaped, 
frequency counts were computed and sent to Dr. Kitzes, NIAMD. 

10. Multiphasic Zone Electrophoresis Programs 

The output from these programs (2000 recipes for Gel Electro- 
phoresis Systems) is being distributed thru NTIS. Our only real 
effort in the past year has been in replacing 3 tapes damaged at 
NTIS and In supplying JCL procedures to the University of 
Colorado to enable them to run the retrieval program, 

11. Radioimmunoassay Programs 

This system of programs which evaluates the Scintillation 
Counter output of radioimmunoassays is in daily use by NICHD. 
In collaboration v/ith Dr. Rodbard several revisions have been 
made to the system to keep pace with changing theory and 
methodology. 

Future plans include the replacement of the plot portion of 
the system with a new printer plot capability. 



9 



66 



12. CBAC/Microf i che Project 

The entire CBAC retrospective file was put onto microfiche 
and the file Inverted with respect to keywords^ molecular 
formulae, and CAS registry numbers for Dr. Fales. 

During the coming fiscal year we plan to create an on-line 
disk file on the PDP-10 and to develop a search program which 
will operate In conjunction with a computer-driven microfiche 
reader. 

13. X-Ray Crystal 1 o^^raohy 

A set of computer programs for implementing direct methods 
In ;i-Ray Crystallography was developed at the Medical Foundation 
of Buffalo. DtIB personnel converted, stored and tested these 
programs for use by Dr. James Sllverton, MHLI. 

Ik. Gastroenterology Data Base - Dr. Krueger, NIAMD 

This will be a data base of journal abstracts from the 
literature of Gastroenterology. During fiscal year 1971+, re- 
format and print programs were written. The CBAC system v/ill 
be utilized for searching the text. 

This will be set up on a recurring production basis with 
a new data tape arriving from the Franklin Institute, Phila- 
delphia, once a month. Volume is expected to be about 3,000 
abstracts/year. The researchers in NIAMD will eventually 
develop and run their own profiles or search requests. 

15. Computerize Data from Scintillation Counter 

In collaboration with Dr. Stephenson of the Viral Carcino- 
genesis Branch, NCI, daca from a Packard Scintillation Counter 
is being transmitted over telephone line., from Hazel tc Labs in 
Virginia to the PDP-10 at NIH. Programs for calculating con- 
centrations of proteins being measured by radioimmunoassay and 
for comparing reactivities of different proteins based on the 
rrlative slopes of their titration curves vyere provided. 
Currently the option of plotting up to 7 distributions on the 
same graph is being Implemented. Future plans Include a test 
for determining the amount of va''iation between different assays 



67 



16. Computerize Data from Coulter Counter 
BJology^ NIAMD. Dr. Shehata, Dr. Kempner 

Coulter Counter data Is supplied to the PDP-10 via paper 
tape from the Laboratory of Physical Biology. Dr. Shehata and 
Dr. Kempner were provided with a program which computes mean, 
mode, standard deviation, and coefficient of variation for 
bacteria counts and displays them at the terminal. Future plans 
include provision of teletype graphs and line-printer graphs. 
Also being discussed are graphic displays, perhaps using the 
Omni graph. 

17. Serum Bank - Inventory and Retrieval System 

Dr. Alexander, Surgery Branch, NCI has data for about 2500 
patients and anticipates up to 10,000 with up to 15 blood 
samples per patient. An Inventory system must be set up to 
keep track of the freezer location of the serum vials and to 
store test results for undefined numbers and types of tests on 
the sera. Also desired are updating and reporting facilities as 
well as a quick retrieval capability. 

Specifications for the system have been developed and 
Implementation has begun. 

18. Mass Spectral Search System 

During FY 197ii, the DMB assumed responsibility for the main- 
tenance of an interactive mass spectral retrieval system. In 
collaboration with Dr. Fales, Laboratory of Chemistry, NHLI, the 
DMB revised all programs In the system to accommodate new file 
formats. A new data file, which Increased the number of spectra 
from 12,000 to 30,000 arrived during the year. This new addition 
plus Interface problems between the system's two computers re- 
quired expert maintenance support. The dynamic nature of this 
system will cause many complex problems for the forseeable 
future. These problems In turn will require thoughtful and 
experienced computer support for solution. 

19. On-Llne Table Creation Facility 

At the request of the Chief, DMB, a procosal for an on-lfne 
table-creation facility has been written. A system of programs 
Is to be developed to do frequency counts and elementary statis- 
tical calculations. It Is to be operated In either a time- 
sharing or batch mode and will permit a user to specify para- 
meters for his job Ir, a problem-oriented notation. 



68 



Math/Stat Section 

1. Analysis of Data of Inbred Mice and Nutritional Treatments. 

in support of DRS programs DMB evaluated data involving 
rations containing two levels of crude protein and three levels 
of fat which were fed to four strains of mice. Split plot de- 
sign and least square analysis programs have been run to evaluate 
the effects of open formula rations on reproductive performance 
and reproductive performance with different physical characteris- 
tics. Evaluation has also included data on reproductive perfor- 
mance and growth of mice fed rations varying in crude protein 
and crude fat. 

2. Interview Scheduling. 

The Clinical Center requested major modifications to 
the original specifications of the Clinical Associate 
Interview Scheduling System. Systems modifications to 
make programs compatible with the new FORTRAN-Gl compiler 
and I/O packages have also been completed. This system is 
continuously being updated to take advantage of new computer 
systems. The original documentation is being modified to 
reflect new changes in the program and run instructions. The 
system Is used each year around April. 

3. Evaluation of CC Heart Data. 

The initial objective of providing a massive storage 
and retrieval capability for chemistry data of the 
Endocrinology Branch, NHLI has been accomplished. Statis- 
tical programs have been prepared to evaluate chemical data 
combined with patient demographic Information and fourler 
analysis on blood pressure data at 2k hour Intervals. The 
final objective of this study will be to develop a capability 
to produce recurring statistical output and evaluations that 
would aid scientists in studying the data. 

k. Retrospective Study of Aortic Valve replacement. 

A set of programs have been completed to edit dat.3 on 

700 patients who underwent open heart operations. Simple 

statistical parameters have aJso been computed in addition to 
survival analyses. 



69 



5. Lipid Protein Study. 

In support of NHLI, this study involves evaiuaticn of 
data on Type 11 kindreds whose propositi were the first 
120 individuals found to have Type M hyp-^r 1 i poprotei nemi a 
at the Clinical Center. To date data description statis- 
tics and multivariate variance analyses have been provided. 
Future plans include the addition of sex and age factors to the 
analysi s. 

5. Probit and Logit. ■' ^ 

This DBS project involves evaluation of different com- 
putational methods in an effort to refine the assessment 
of biologies products. The programming has undergone major 
modifications; it computes probits and logits with options to 
compute individual or pooled slopes, with or without con- 
version of input data. Additional options in the programs 
include tests for parallelism and relative potency. 

7. Evaluation of Pertussis Potency Tests. * 

This project involves the evaluation of Pertussis 
vaccine potency as performed at DBS. Programs performing 
probit analysis have been extensively utilized in this 
study. However, additional programming is underway to 
determine if there are seasonal variations in the test as 
evidenced by cyclic changes in standard vaccines or changes 
in the immunity of animals receiving standard vaccines. The 
findings so far have led to two major innovations in procedure, 
an increase in the number of animals per dose and careful 
ranflomi zat ion of mice to vaccines and dose levels. 

8. Multiple Family Group Study. 

The basic unit of this study is a group therapy sessic,= in 
which several families with disturbed adolescents meet with 
personnel from the NIMfl Adult Psychiatr> Branch and the Psychiatic 
Institute. There have been approximately 100 of these sessions, 
each session yielding five types of data: 



Cohesi/eness Questionnaire Data. 
"Who-to-i/hom" Speech Data. 
V^ard Value Questionnaire Data, 
Sociometry Data. 
Seating Position Data. 



The first three sets of data have reached the analysis stage. 
The sociometry data is still in the process of being validated 
and nothing has been attempted with the seating position data. 



70 



9. Type II Intervention Study 

This effort is part of the analysis phase of the Type II 
Study being conducted by the NflLl. In collaboration with Dr. 
Brensil<e and the MHLI Biometrics Branch, four PL/1 data retrieval 
programs, initially begun by user, have been implemented. Upon 
execution, these programs retrieve data pairs for each patient; 
eliminating missing or inadmissible data and input these data to 
a program that computes paired-t tests. 

A program has been developed to retrieve and edit different 
sets of data and to access the SPSS (Statistical Programs for 
Social Studies) package. One of the features of this package 
will provide Dr. Brensike with the flexibility to obtain what- 
ever cross-tabulations he wishes. 

Future plans include the Implementation of the Duncan- 
Walker probability package for analyzing independent variables. 

10. Treatment of Massive Obesity: Changes In Cardiovascular 
Risk Factors 

This study involves analysis of changes in cardiovascular 
risk factors (I.e. blood pressure, glucose tolerance, blood 
lipid levels, uric acid, heart size on chest X-ray, electro- 
cardiographic findings) associated with weight loss in patients 
with massive obesity. One hundred six patients who lost at 
least one hundred pounds constitute the study population. 

The degree of weight loss and the changes in the above risk 
factors assessed by seventeen parameters measured before and 
after weight loss are being subjected to systematic analyses In 
terns of incidence of and degree of abnormality. Changes are 
correlated with degree of weight loss. Correlations among and 
between parameters are oeing der'ved and analyzed. 

11. Carcinogenesis Information Evaluation 

The analysis phase of this study involves data from ongoing 
bioassay experiments In mice. The immediate aim is to detect 
previously unidentified carcinogens. 

A major programming effort has begun on this project. Data 
for the statistical analysis v/ill be drawn from two files: one, 
a file containing complete background data for each animal 
group; the second, a file containing complete Individual animal 
records. Call these the 'Control File' and the 'Individual 
File' respectively. Initial analysis programs Include simple 
data descriptions, such as counts of animals within a sample 
and counts of animals with certain pathologies. Survival 
statistics are also being developed. 



71 



12. Radioimmunoassay of GS Antigens 

Project involves development of a system to calculate the 
degree of antigen bound in radioimmunoassay data using one^ two 
or three isotopes. In this procedure the amount of antigen Is 
measured by inhibition of the precipitation of standard labelled 
antigen and specific antibody. The purpose of the second and 
third Isotopes Is to determine non-specific contributions to 
the precipitate. 

Programs have been completed to perform calculations using 
one and two isotopes and the Investigator has received numerous 
outputs. Programming for three isotopes has just begun. 

13. Analysis of Data In Occupational Medicine 

Project involves analysis of medical examination data and 
dynamic ECG records of NASA Headquarters personnel. The Initial 
phase of this project involving regression analysis Is complete. 
No future assistance is planned for this effort. 

lU. Rvaluatlon of Long Lines models 

This project involved the conversion and evaluation of three 
separate systems that compute econometric models used to derive 
AT&T long lines rates. All of the programs were written by AT&T. 
DMB's objective In this project was to assist FCC in evaluating 
these models. The system Is now operating on the DCRT computer 
and has been turned over to FCC personnel. 



72 ■ ^ 14- 



NIH 

l|^ Amazing Resi 



Research. 

Amazing Help. 



http://nihlibrary.nih.gov 



10 Center Drive 

Bethesda, MD 20892-1 150 

301-496-1080 




3 1496 00821 8664 



12