I’m interested in creating a new cortical parcellation based on my own geodesic distance-based algorithm. For instance, I would like to extract the basins area of the cortex. I am wondering if it is plausible to directly generate the parcellation on a template image, such as the fsaverage lh.pial files. Has anyone had any experience with this or have any insights on whether this approach would be suitable? Any guidance or suggestions would be greatly appreciated.