Hi all, I¹m really stuck on a problem I was hoping to get help with.

I¹m having trouble determining the correspondence between ROIs in a surface-based annotation and those in a volume-based annotation generated using mri_aparc2aseg. I have some custom-generated surface annotations and have been using a recently updated binary of mri_aparc2aseg which adds 10,000 to the values of left hemi annotation labels and 20.000 to the right hemi labels in the converted volume, to help distinguish the two hemisphere¹s rois.

To check the correspondence between the surface and volume, I chose an roi 10386 from the volume. To check it on the surface I did the following in matlab:

[v,l,ct]=read_anntotation('lh.fsavg_500_compact.aparc.annot'); new=zeros(size(v)); new(find(l==ct.table(386,5)))=1; % note the roi index should be 10,000 less that of the volume. F=write_wfile('386.w',new,v);

I did this for both the fsaverage surface and a single subject surface, and compared it with the volume representation for the same single subject. The attached pic called vol386 shows the roi in blue, overlaid on the volume parcellation in greyscale. It is located in the left inferior temporal lobe.

The attached surf386 pic however, indicates this region is in the retrosplenial cortex for both the single subject (left) and fsaverage (right). Am I missing something here? Have I misunderstood the correspondence between surface and volume rois?

Thanks for your help, Alex