Summary: All tracer-kinetic models currently employed with positron-emission tomography (PET) are based on compartmental assumptions. Our first indication that a compartmental model might suffer from severe limita tions in certain circumstances when used with PET oc curred when we implemented the Kety tissue-autoradiog raphy technique for measuring CBF and observed that the resulting CBF estimates, rather than remaining con stant (to within predictable statistical uncertainty) as ex pected, fe ll with increasing scan duration T when T > 1 min . After ruling out other explanations, we concluded that a one-compartment model does not possess suffi cient realism for adequately describing the movement of labeled water in brain. This article recounts our search for more realistic substitute models. We give our deriva tions and results for the residue-detection impulse re sponses for unit capillary-tissue systems of our two can didate distributed-parameter models. In a sequence of trials beginning with the simplest, we tested fo ur progres-The measurement of CBF is important not only because of the information it provides about the normal and diseased brain, but also because it is essential to other important measurements (e.g. , rate of oxygen consumption). Tr acer-kinetic models are employed with external radiation-detection de vices to interpret dynamic imaging data in terms of the movements of radiolabeled water (H 2 1 50) or other diffusible tracer (e.g. , [ 1 8F] fluoromethane) for the purpose of inferring CBF. The tracer-kinetic model currently employed with positron-emission tomography (PET) for this purpose is based on
443sively more detailed candidate models against data fr om appropriate residue-detection experiments. In these, we generated high-temporal-resolution counting-rate data re flecting the history of radiolabeled-water uptake and washout in the brains of rhesus monkeys. We describe our treatment of the data to yield model-independent em pirical values of CBF and of other parameters. By substi tuting these into our trial-model fu nctions , we were able to make direct comparisons of the model predictions with the experimental dynamic counting-rate histories, con firming that our reservations concerning the one-com partment model were well founded and obliging us to re ject two others . We conclude that a two-barrier distrib uted-parameter model has the potential of serving as a substitute for the Kety model in PET measurements of CBF in patients, especially when scan durations for T> 1 min are desired. Key Words: Cerebral blood flow-Dis tributed-parameter models-Po sitron-emission tomog raphy-Tissue heterogeneity-Tr acer kinetics. compartmental assumptions; that is , tracer is as sumed to move between discrete subvolumes, or "compartments," within each of which tracer is assumed to distribute instantaneously upon arrival. Thus, in a compartmental model, gradients of con centration are assumed to be zero (i.e. , their spatial profiles flat) within each compartment at all times. The compartmental mode...