Hi Burak,

Your best bet is to keep this question going to the list.  I’ll leave it for Bruce, or Doug, or Nick to answer.

Jenni

---

From: Burak Ozkalayci [mailto:burak.ozkalayci@ugent.be]
Sent: Thursday, April 19, 2007 7:11 AM
To: Jenni Pacheco
Subject: Re: [Freesurfer] Grainy thickness ribbon

Hi Jenni,

In fact no. I need a ribbon whose voxel values are assigned according to the cortical thickness. the mri/?h.ribbon.mgz is just a binary mask for the ribbon (as you can check from the cortical thickness ribbon images I attached to my former mail). I think that the -fillribbon option for the mri_surf2vol function does not work properly when the pial surface is selected like:

mri_surf2vol --volreg reg.dat --surfval ../surf/lh.thickness_max --hemi lh --template ./orig/001.mgz --outvol leftThickness_max_filled_pial.nii --fillribbon --surf pial

the ribbon is filled in the normal directions of the surface which seem to give correct results for the white surface but for pial surface the direction of the normals should be reversed I think. By comparing the images I attached to the former mail, you can see that the filled ribbon from the pial and the white surface is different.

I hope I could make it clear.

Thanks again for any response, help.

Burak



Jenni Pacheco wrote:

Hi Burak,
 
I'm not really sure what you are trying to do, but a cortical ribbon is
generated as part of the recon process, mri/?h.ribbon.mgz.  Perhaps this is
what you are looking for?
 
Jenni
 
-----Original Message-----
From: freesurfer-bounces@nmr.mgh.harvard.edu
[mailto:freesurfer-bounces@nmr.mgh.harvard.edu] On Behalf Of burak ozkalayci
Sent: Wednesday, April 18, 2007 6:39 AM
To: Freesurfer Mailing List
Subject: [Freesurfer] Grainy thickness ribbon
 
Dear surfers,
 
I try to get a thickness map of the cortex by using the  mri_surf2vol 
function. However when I used it like :
 
 mri_surf2vol --volreg reg.dat --surfval ../surf/lh.thickness_max --hemi 
lh --template ./orig/001.mgz --outvol leftThickness_max_filled_pial.nii 
--fillribbon
 
it generates a very grainy map. I think that those unassigned voxels are 
not hit by the normals of vertices on the white surface. To  compensate 
it I tried to fill the ribon with the normals originated from the pial 
surface by :
 
 mri_surf2vol --volreg reg.dat --surfval ../surf/lh.thickness_max --hemi 
lh --template ./orig/001.mgz --outvol leftThickness_max_filled_pial.nii 
--fillribbon --surf pial
 
however the ribbon I get is shifted to the outer side of the cortex. I 
think the normals of the pial are oriented to show outer space of the 
cortex and they hit the voxels at the outerspace. I also attached the 
thickness maps of the ribbon I get from the white and the pial surfaces 
and the T1 image. They are the same slice of the volume, and you can see 
that pial one passes beyond the pial surface in general.
 
In the help of mri_surf2vol for the fillribbon option, it is explained 
as "iterate projfrac from 0 to 1". And when the code is running I saw 
that iteration has a  0.05 step. I wonder whether is it possible to 
define the stepsize and the range for example 0 to -1 to reverse the 
hitting voxels.
 
Thanks in advance.
 
Burak