Regional Quantitative Analysis of Cortical Surface Maps of FDG PET Images

Abstract

Cortical surface maps are advantageous for visualizing the 3D profile of cortical gray matter development and atrophy, and for integrating structural and functional images. In addition, cortical surface maps for PET data, when analyzed in conjunction with structural MRI data allow us to investigate, and correct for, partial volume effects. Here we compared quantitative regional PET values based on a 3D cortical surface modeling approach with values obtained directly from the 3D FDG PET images in various atlas-defined regions of interest (ROIs; temporal, parietal, frontal, and occipital lobes). FDG PET and 3D MR (SPGR) images were obtained and aligned to ICBM space for 15 normal subjects. Each image was further elastically warped in 2D parameter space of the cortical surface, to align major cortical sulci. For each point within a 15 mm distance of the cortex, the value of the PET intensity was averaged to give a cortical surface map of FDG uptake. The average PET values on the cortical surface map were calculated for each cortical point in a lobe, and were compared to those obtained by averaging across the entire ROIs on the 3D PET image directly. The average regional values obtained from the cortical surface map were found to match well with those taken directly from the 3D PET images. The results demonstrate that cortical surface maps of FDG PET images give accurate quantitative information on regional brain function, in addition to offering an improved visualization.