I am wondering because when I look at the segmentation for fsaverage, it seems like it is missing a lot of the gray matter, particularly in the big sulci, most likely from the 'smoothing' that comes with averaging subjects together. And therefore I wonder if this is the best thing to use as an intermediate step. Are you instead suggesting (as it seemed like in this thread), to put the SPM roi into individual subject space and then somehow project it to the surface in free surfer?