Hi guys,
Happy New Year, all.
When I try to adwarp a functional dataset in subject-space to its TT-normalized anatomical volume, adwarp starts generating a massive dataset which I have to cancel. This is a known issue, and requires mapping a voxel from a spatially normalized activation map back to a corresponding location in the subject space on a per-voxel basis with adwarp. It works, but its tedious and not robust.
Instead of doing that, I discovered the joys of FreeSurfer/SUMA mapping, and its amazing for cortical stuff.
However, as far as I know, FS can't parcellate and construct surfaces for basal ganglia and deep nuclei. If this is true, then tlrc'ing is the only option for these regions, right?
If so, is there any way to get around the problem of having insanely large spatially normalized functional datasets?
-Prantik