I am trying to automate ROI creation in freesurfer following the VolumeRoiCorticalThickness tutorial, based on the AAL defined ROI-s from MarsBar.

QUESTION

ROI5.nii was originally in the same space as TT_avg152T1.nii (see MORE INFO below).

I understand that after Step 2  (when  TT_avg152T1.nii is registered to fsaverage with spmregister), TT_avg152T1 will be in fsaverage space. 

But, we didn't transform or register ROI5.nii, so I understand that it will continue in the same original space. 

When I visualize ROI5 over the original TT_avg152T1 it is ok, but when I visualize it over the fsaveraged one, it is not. 

Shouldn't we need to register or normalize ROI5.nii to fsaverage as well prior to the mri_vol2surface mapping?

MORE INFO:

1.- I created the ROI5.nii in MarsBar: "ROI definition" > "Export" > "Export ROI to image" (select: MNI_Frontal_Inf_Tri_L_roi.mat) > from image (select: avg152T1.nii) > Save it (name: ROI45.nii)

2.- I follow exactly the same steps as the tutorial, but in step 3) when I check the ROI in the inflated surface, it is clearly out of place...