A MULTI-STATION COMPUTER SYSTEM FOR 
ACQUIRING AND PROCESSING ANIMAL DATA 
Eugene F. Uretz, John A. McClary and Steve Morgan* 
This paper describes a system used to acquire both 
laboratory and physiological data under control of 
users at terminals distributed throughout the T & E Fa- 
cility of the Medical Devices Applications Program of 
the National Heart and Lung Institute. The facility is 
operated by the IIT Research Institute in Chicago. The 
structure of the system allows data from many sources 
to be acquired simultaneously with complete integrity 
at rates up to 500 samples per second and associates 
contextual description with all acquired data. These de- 
scriptions uniquely identify a given data series by de- 
scribing the test under which it is acquired; the animal, 
device and/or sample associated with the data; and the 
time of execution of the test. The features of the system 
which allow acquisition and processing to proceed si- 
multaneously without delay to the terminal user are de- 
scribed. One feature is a multi-level priority structure, 
which allows high priority tasks such as data sampling 
to take precedence over lower priority tasks such as file 
searching. Another is a unique contextual dayfile struc- 
ture which allows all data from all terminals to be 
mixed together in a single file while preserving the con- 
textual identifiers associated with the data. The system 
also allows automatic storage of high rate physiological 
data on the computer disk at high priority, where it can 
be analyzed and processed at low priority without com- 
promising the integrity of the system. Another feature 
is the assignment of priorities to users' jobs so that short 
computational jobs are not delayed by slower jobs. The 
data acquired or derived by the system is transferred 
once a day to a large batch processing system via data- 
phone where it is stored in organized files. Many 
months of data are held in these files where they are 
used for further numerical and statistical analysis, as 
well as for report generation. Several examples of spe- 
cific applications of the system are described. These in- 
clude manual logging of biochemical data, automatic 
read in and processing of spectrophotometer, densitom- 
eter, and coulter counter data, and Fourier analysis of 
blood flow and pressure. 
INTRODUCTION 
The system described is part of the capability 
developed to meet the computational needs of 
♦ IIT Research Institute, Test and Evaluation Tacility (sponsored 
by MDAP, NHLI) Chicago, Illinois. 
IITRI operated T & E Facility of the Medical 
Devices Applications Program (MDAP) of the 
National Heart and Lung Institute (NHLI). 
The T & E Facility's job is the test and evalua- 
tion of medical devices. The devices are varied 
and the tests quite comprehensive including in- 
vivo and in-vitro tests v^^hich may be medical, 
biological, mechanical, electronic, etc. Specific 
tests vary markedly according to the nature of 
the device being evaluated. Furthermore, a sin- 
gle experiment may involve the collection of a 
great deal of data on many animals. Thus, eval- 
uation of the Dow oxygenator involved running 
total or partial bypasses on many animals, ob- 
taining many biochemical data values on each 
animal over a period of days, and statistically 
evaluating and correlating all data from all ani- 
mals. 
The implication in developing the T & E Fa- 
cility Computational capability was clear : To be 
effective, the system developed had to be able to 
reliably acquire, store and manipulate large 
amounts of data of a diverse nature. To keep 
the reliability high, it was important to aid the 
lab worker in the acquisition and analysis of 
data of all types. 
The overall system developed to satisfy these 
computational capabilities is a two-tiered sys- 
tem (Figure 1). The first tier is a small real 
time system used to aid laboratory and scientific 
workers in performing experiments, in acquir- 
ing and analyzing data, and in placing suitable 
identifiers with all data. It can store up to one 
full day's data with their identifiers. 
Another very important function performed 
by the first tier system is the periodic transmis- 
sion of all data and their identifiers to the re- 
mote second tier system. The collected data can 
thus be entered into the appropriate second tier 
1243 
