mri_vol2surf goes through two steps to map the functional to fsaverage. First, it maps to the subject's surface (it knows what this is because it has the subject's ID in the register.dat file). It then maps the subject's surface to fsaverage throught the sphere.reg created during recon-all. Again, it knows where to find this from the subject's ID.
doug
On Fri, 8 Dec 2006, Emily Cooper wrote:
hi,
i have been using mri_vol2surf to map individual functional data onto the fsaverage brain:
mri_vol2surf --src 003/bold/03.time.intval.resam+orig.BRIK --src_type afni --srcreg 003/bold/register.dat --hemi rh --projfrac-avg 0 1 10 --trgsubject fsaverage --out 003/surf/03.time.intval.fsave.mgh
each time i do this, i compare the mapping onto the curvature with the same functional mapped onto the individual brain:
mri_vol2surf --src 003/bold/03.time.intval.resam+orig.BRIK --src_type afni --srcreg 003/bold/register.dat --hemi rh --projfrac-avg 0 1 10 --out 003/surf/03.time.intval.orig.mgh --out_type mgh
and thus far it's always been a fantastic match. i was curious, how does the mri_vol2surf script match up the individual functionals onto the curvature of fsaverage? i had a theory that maybe it implicitly references the surf directory of the subject, so i made a new subject directory tree with the functionals of one subject (003) and the surf directory of another. vol2surf-ing the same functional in this new directory to fsaverage results in an identical mapping as to in the regular 003 directory. a similar SUMA function (mapicosahedron) maps functionals into an icosahedron space by first warping the individual anatomical. does freesurfer's mri_vol2surf work somehow without using this information?
thanks, emily