Reproducibility of the brain perfusion index for measuring cerebral blood flow using technetium-99m compounds

We previously developed an alternative method for estimating the brain perfusion index (BPI) using technetium-99m compounds and spectral analysis (SA) for quantification of cerebral blood flow (CBF). In this study, we investigated the reproducibility of the BPI values obtained by SA (BPI^S) using a double injection of technetium-99m ethyl cysteinate dimer without any intervention, and compared it with that of the BPI values obtained by graphical analysis (BPI^G). The BPI^S values in the first (x) and second sessions (y) correlated closely (y=0.921x+0.036; r=0.962; n=64; s.e.e.=0.058 min^-1). Although the BPI^G values in the first (x) and second sessions (y) also correlated (y=0.942x+0.040; r=0.916; n=64; s.e.e.=0.061 min^-1), the correlation coefficient for BPI^S was significantly higher than that for BPI^G. The reproducibility was dependent on the injection dose ratio of the second session to the first (R). The difference in BPI between the first and second sessions tended to be smaller when 1.5<R<2.5 than when 0<R<1.5 or 2.5<R. These results suggest that the reproducibility of BPI^S is satisfactory and that it can be applied to two sequential measurements of CBF using a double injection of ^99mTc compounds. When applying our method to such measurements, it is recommended that R is taken as approximately 2.0.