1222 
MONITORING 
nels are sampled. It is then the responsibility of 
the CLKINT routine to sort out the information 
of only those channels requested by the users. 
Sampling is done at a minimum period of five 
msec, 200 samples per second, and samples can 
be obtained at any interval which is a multiple 
of the five msec period up to periods of several 
seconds. 
DISKIO: This subroutine is responsible for 
handling the transfer of all data betvi^een the 
user programs and the fixed-head disk. The 
data is stored under a file structure using four 
diiferent sort keys: Patient Number, Device 
Number (the number of the device under test) , 
the Program Number of the program which 
generated the data, and the Time at which the 
data was obtained. The disk is divided into 
pages of 180 words each. DISKIO is responsible 
for sorting and writing the data on the correct 
pages as well as reading the correct data when 
any request is made for information based on 
the four sort keys. If data from a new patient is 
to be added to the disk, this subroutine finds 
those pages which are available and stores the 
data. 
Since the datafile is maintained by chaining 
pages together according to the two sort keys, 
Patient Number and Device Number, it is the 
responsibility of the subroutine to maintain this 
chaining. Also, if a patient is to be deleted, DIS- 
KIO stores the necessary information for book- 
keeping and returns the pages used by that pa- 
tient to the "Page Available" pool. Variable 
length records can be stored, but the user pro- 
gram must specify the length of the record 
which it is generating. It can be seen that since 
DISKIO is the only way in which a user pro- 
gram can write data on the disk, it is totally re- 
sponsible for the maintenance of the datafile 
and will greatly reduce the probability of data 
being destroyed because of incorrect informa- 
tion which might come from the user programs. 
Remote Terminals 
To use the computer system, remote termi- 
nals shown in Figure 4 were selected and are lo- 
cated at various sites throughout the Facility. 
This terminal uses a storage tube for its display 
and was selected because it has a 512 by 512- 
point graphic capability and also can display 
Alpha-numeric data. It is also relatively low in 
cost as compared to other terminals with the 
same graphic capability. The terminal interacts 
with the computer by using two octal switches 
and a 12-key keyboard. Located between the 
two octal switches are two call buttons, one for 
each octal switch, an indicator light which illu- 
minates when the station has been recognized 
by the computer and is active, and a manual 
button for erasing the oscilloscope. Another fea- 
ture which is available on the terminal is that 
of being able to display analog signals. Three 
signals can be connected to the back of the ter- 
minal, and by pressing the analog display but- 
ton, the three signals are displayed simultane- 
ously. The sweep rate can also be varied. If 
information comes from the computer during the 
period in which the analog signals are being 
displayed, the display is automatically termi- 
nated and the information from the computer is 
written on the oscilloscope. 
PROGRAMS 
When a program is to be run from a remote 
terminal, the program number is dialed into the 
octal switch and the call button depressed. This 
interrupts the computer, which then reads the 
octal switch, determines which program is to be 
used, retrieves the program from the disk, and 
stores it in one of the available user areas. Con- 
trol is then given to the program and a message 
is written on the oscilloscope indicating which 
options are available in the program and how to 
proceed with its operation. 
Dye Dilution Curve 
One of the user programs which is currently 
available on the system does analysis of indica- 
tor dilution curves. It is described in some de- 
tail to illustrate how the programs are de- 
signed. Using the Stewart-Hamilton method, 
this program calculates cardiac output and 
other parameters which are of hemodynamic 
importance, such as, appearance time, build-up 
time, mean circulation time, and central blood 
volume. If this were the program called, the dis- 
plays that would occur on the oscilloscope are 
shown in Figure 5. The first display is the iden- 
