1160 
MONITORING 
estimate of the time constant can lead to large 
errors in the calculated area. I think that you 
can readily appreciate the uncertainties inherent 
in this approach. 
Now, our modified recirculation model says 
that in order to compute the mean transit time, 
you have to make an additional injection which 
gives another response curve you can call qv(t). 
This curve represents the actual quantity of 
tracer in the detector field of view due to the 
venous injection. The modified formula reads 
like this: [Equation (10)]. Here, q.^_ is the 
steady-state value that the venous and arterial 
curves approach. If each is normalized to the 
amount injected, they will both approach the 
common value q^. . As you can see from this 
formula [Equation (10)], the mean transit 
time on our model is related not to the area 
tinder the arterial curve, but to the area bettveen 
it and the venous curve. Then, of course, by the 
central-volume principle, one can compute rela- 
tive blood flow [Equation (9)]. 
Chairman Warner: I don't really under- 
stand the basis for your saying that you're not 
dependent upon getting good mixing though. 
You're measuring instantaneous differences in 
concentration in two locations and . . . 
Dr. Larson : I think what I said was that our 
model makes no assumptions as to tracer-trans- 
port mechanisms within extravascular tissue. 
Our equations are derived on the basis of tracer 
conservation alone. They say nothing at all 
about how tracers diffuse within the tissues. 
Chairman : No, but you're going to have to. 
In order to deduce what is happening to the 
blood, you've got to know how the tracers are 
distributed in the blood, don't you? 
Dr. Larson : I'm sorry, Mr. Chairman ; I did i 
not realize you were referring in your previous ! 
question to mixing in the blood. One of the j 
assumptions of the central-volume principle 
[Equation (9)] is that there is complete mixing , 
of tracer with the blood at the arterial inflow | 
of the vasculature of interest. However, in order i 
to compute the mean transit time on our model, I 
we need no assumptions about the degree of i 
tracer mixing anywhere else, whether inside or [ 
outside the detector field of view. I 
Questioner (Unidentified) : I take it that 
these two injections are made at two different 
times. 
Dr. Larson : Yes, they can be. In our experi- | 
ments, we have made them sequentially. I sup- | 
pose there is no reason they couldn't be made I 
simultaneously if you had two channels for f 
resolving the response curves. [ 
Questioner: Two different isotopes? ; 
Dr. Larson : Yes — two different radioisotopes f 
labeling the same tracer substance. [ 
D. B. Jackson, Abbot Laboratories, North | 
Chicago : Does X.V/F divided by the mean transit j 
time equal one? f 
Dr. Larson: Yes, because by the central- 
volume principle, the mean transit time is equal [ 
to XV/F. ( 
Dr. Jackson: No, what's the XV mean? f 
Dr. Larson : X is the average partition coef- | 
ficient. We define this to be the ratio of the con- | 
centration of the tracer in the tissue to the I 
concentration in the blood. j 
Dr. Jackson: What is V? | 
Dr. Larson: V is the geometrical volume in 
which the tracer is distributed. 
