Correction for partial-volume effects on brain perfusion SPECT in healthy men.

UNLABELLED The limited spatial resolution of SPECT scanners does not allow an exact measurement of the local radiotracer concentration in brain tissue because partial-volume effects (PVEs) underestimate concentration in small structures of the brain. The aim of this study was to determine which brain structures show greater influence of PVEs in SPECT studies on healthy volunteers and to investigate aging effects on SPECT after the PVE correction. METHODS Brain perfusion SPECT using (99m)Tc-ethylcysteinate dimer was performed in 52 healthy men, 18-86 y old. The regional cerebral blood flow (rCBF) was noninvasively measured using graphical analysis. SPECT images were corrected for PVEs using gray-matter volume, which was segmented from coregistered MR images and convoluted with spatial resolution of SPECT scanners. Absolute rCBF data were measured using a 3-dimensional (3D) stereotactic template for regions of interest on anatomically standardized SPECT. We examined correlation of advancing age with rCBF before and after the PVE correction. To validate the correction method for PVEs, a Hoffman 3D brain phantom experiment was also performed. RESULTS The PVE correction remarkably reduced the coefficient of variation for SPECT counts in the whole phantom. The PVE correction made the rCBF distribution more homogeneous throughout the brain with less intersubject variation than the original distribution. There were significant negative correlations between age and adjusted rCBF in the bilateral perisylvian and medial frontal areas. These correlations remained significant after the PVE correction. Instead of a positive correlation in the medial temporal structures between age and adjusted rCBF before the PVE correction, the sensorimotor and parietal areas mainly showed positive correlations after the correction. CONCLUSION SPECT data reflect both brain volume loss and functional changes. Use of the PVE correction in brain perfusion SPECT provides a more accurate determination of rCBF even in healthy volunteers.