Click on image to view larger version.
FIGURE 1. Procedures of correction algorithm for PVEs in SPECT data using MR images. Original SPECT image (B) was coregistered to MRI after scalp editing (D) of original MR image (A) using binary mask for whole brain (C). Original MRI was segmented into white matter (F) and gray matter (G). Segmented images were convoluted with point spread function of SPECT device. White-matter SPECT image (H) was then simulated from convoluted white-matter MR image (H). For this simulation, maximum count for white matter of SPECT, ROI was automatically determined by setting threshold to >95% of maximum count density of white-matter MR images (J). Gray-matter SPECT images (K) were obtained by subtraction of simulated white-matter SPECT image from original SPECT image coregistered to MRI (E). Lastly, gray-matter SPECT image (K) was divided by convoluted gray-matter MR image (I) on voxel-by-voxel basis. Gray-matter SPECT image with PVE correction (M) was finally obtained from application of binary mask for gray matter (L) to divided image.
