Douglas,

The previous message had an images attached that seemed to be too big. it was "bounced" automatically. I guess too many images were attached. 

I ran the command for the rh.white and lh.white and the image is not perfectly registered. I tried a lot of things, but nothing worked. I ended up using the idea from Bruce. I Binarized the aseg+aparc file, extracting the wm regions, eroded, and subtracted. That gave me a surface as a volumen. This also can be done with only the images (no need to extract the whole freesurfer output)  The image from mri_surf2vol had better quality, but I couldn't get it to register. If by any chance someone figures this registration problem out, I will use certainly use it. 

Thanks for all your help. 

Gabriel 


previous message: 

Douglas, 

I extracted the whole directory and ran:

mri_surf2vol --mkmask --surf white --hemi rh --o rh.white.mgz --template ../subjects/subject/mri/T1.mgz  --sd ../subjects --subject subject --identity subject
gdiagno = -1
Using identity matrix for registration
Overriding reg subject subject with subject
Qa2v: SurfXYZ to VolCRS: ------------------------------
-1.000   0.000   0.000   128.000;
 0.000   0.000  -1.000   128.000;
 0.000   1.000   0.000   128.000;
 0.000   0.000   0.000   1.000;
--------------------------------------------------
subjects dir   ../subjects
hemi           rh
mksurfmask     1
projfrac       0
outvol   path  rh.white.mgz
template path  ../subjects/subject/mri/T1.mgz
------- Anat2Vol Registration (TkReg)----
 1.000   0.000   0.000   0.000;
 0.000   1.000   0.000   0.000;
 0.000   0.000   1.000   0.000;
 0.000   0.000   0.000   1.000;
-----------------------------------------
height = 256
 width = 256
 depth = 256
 xsize = 1.000000
 ysize = 1.000000
 zsize = 1.000000
  cdc  = -1.000000 0.000000 0.000000
  rdc  = 0.000000 0.000000 -1.000000
  sdc  = 0.000000 1.000000 0.000000
  xyz0 = 3.373505 20.515587 -6.721008
Gdiag_no  -1
Reading surface ../subjects/subject/surf/rh.white
Done reading source surface
surf nframes = 1
INFO: mapping vertices to closest voxel
INFO: resampling surface to volume
INFO: sampled 84160 voxels in the volume
INFO: writing output volume to rh.white.mgz
done

I did both hemispheres, they are not registered. I tried using as template the T1.mgz, and T1 converted to nifti. I also tried output as mgz and then mri_convert to nifti, the also output to nifti. I always obtained the same volumes, and they were always unregistered.


both hemispheres, they are not registered. I tried using as template the T1.mgz, and T1 converted to nifti. I also tried output as mgz and then mri_convert to nifti, the also output to nifti. 

Inline image 2