[Estimation of integral of input function for quantification of cerebral blood flow with...

The present study was designed to investigate a possibility of substitution of the venous blood radioactivity counts sampled 26 min post injection for the octanol-extracted arterial blood radioactivity counts obtained at 5 min after the injection of N-isopropyl-p-[123I]iodoamphetamine (123I-IMP). Furthermore, we investigated whether the integral of input function can be estimated from the venous blood radioactivity counts sampled 26 min post injection and the whole-brain time-activity curves early after 123I-IMP injection. There was a good correlation between the arterial blood radioactivity counts sampled 5 min post injection (y) and those obtained at 26 min (r = 0.902; n = 91; y = 2.348x - 867.063). There was also a good correlation between the arterial (x) and venous blood radioactivity counts (y) sampled 26 min post injection (r = 0.954; n = 14; y = 0.761x + 924.336). The venous blood radioactivity counts sampled at 26 min (x) correlated well with the octanol-extracted arterial blood radioactivity counts sampled at 5 min (y) (r = 0.964; n = 32; y = 0.173x - 21.598). There was a good correlation between the integrals of input function obtained from the regression equation obtained above and the whole-brain time-activity curves acquired during 7 min post injection (y) and those obtained by 5-min continuous arterial blood sampling (x) (r = 0.965; n = 41; y = 0.957x + 2665.208). These results indicate that this noninvasive and simple method can estimate the integral of input function for quantification of cerebral blood flow using 123I-IMP.