Dear FreeSurfer experts,

I have a question on the usage of mri_vol2surf's option, "--projopt". What I want to do is to map the diffusion scalar value (like mean diffusivity [MD]) to the cortical surface. Though mri_vol2surf with --projfrac 0.5 worked pretty well, I also tried with --projopt which I assumed to yield better dealing with the partial volume effects.

However, when I run with --projopt, I encountered the following errors. 

.......

error: GM estimation at vno 31257 couldn't find enough samples
error: GM estimation at vno 31258 couldn't find enough samples

.......

It looks the code tells me there is not enough data to solve the problem.

Could you please instruct me what should I do to use mri_vol2surf with --projopt ?

My diffusion maps are in 1.7mm isotropic voxel and aligned with anatomical T1w with BBregister, and I used the code as below.

################

tkregister2 --mov ${SUBJECTS_DIR}/${subject}/mri/norm.mgz --s ${subject} --regheader --reg regheader.subject.dat --noedit

mri_compute_volume_fractions regheader.subject.dat ${SUBJECTS_DIR}/${subject}/mri/norm.mgz pvf 

  # this gave me pvf.wm.mgz, pvf.csf.mgz, pvf.cortex.mgz, which are in 1mm voxel

mv pvf.cortex.mgz pvf.gm.mgz

  # rename to pvf.gm.mgz to be compatible with the file name that appears in mri_vol2surf code

mri_vol2surf --mov MD.nii.gz --regheader ${subject} --hemi lh --surf white --projopt pvf --o lh.MD_pvf.GM.mgz

################

Respectfully yours,