Modeling strategy for background compensation in repeat cerebral blood flow quantitation with diffusible tracers

A method has been developed for repeat quantitation of regional cerebral blood flow (rCBF) using a diffusible tracer such as O-15 water (H/sub 2//sup 15/O) and I-123 iodo-amphetamine (/sup 123/IMP). The background activity distribution due to the previous tracer administration was formulated according to a compartment model, so as to allow initiation of the next scan while residual radioactivity exists, minimizing the intervals between scans. This background distribution was estimated from the previous scan with the minimum enhancement of the statistical noise. Theoretical simulation study was carried out, and demonstrated that the estimated images of the transient tracer distribution are highly weighted on transient rCBF immediately after the tracer administration, and that the change of rCBF after a certain period has only small contribution to the estimated tracer distributions. Another set of simulation study showed that estimated rCBF was sensitive to the transient rCBF only immediately after the tracer administration, and that the change after a certain period does not contribute to the rCBF estimated by the present approach. This technique was then applied to a clinical test-retest study using IMP and SPECT. Reproducibility of rCBF quantified by this approach was less than 5% in 5 studies performed at rest-rest condition, and vascular reactivity determined by a rest-vasodilating drug stimulation (Rest-Diamox) on 5 healthy volunteers appeared to be 40%, which was in an agreement with literature value. We thus conclude that repeat rCBF quantitation is feasible with considerably shorter intervals than the present protocol, and can be applied to clinical studies.