Hi Rafa
Might be because MATLAB uses 1-based indexing?
What if you compare the values of the PET in tksurfer and vol(numberOfVertex + 1) ?
Best Krish
On Sep 3, 2010, at 10:17 AM, Rafa x wrote:
Dear Freesurfers, I’m trying to obtain in matlab the PET value of each vertex of the cortex from one freesurfer subject.
To calculate it, first I do the coregister with: bbregister -–mov PETimage.nii –-t2 –-s subjectName –-init-fsl –-reg register.dat
I have checked the result and it is provides perfect corregister between the PET and MRI. Then I calculate the PET in the cortical surface with: mri_vol2surf -–mov PETimage.nii –-reg register.dat –-hemi ?h –projfrac 0.5 –-interp nearest –-o PETSurface.mgh
I load it with tksurfer like an overlay. It shows the variability of the PET activity across the cortex without problem. At last I extract the parameters in Matlab with: [vol, M, mr_parms, volsz]=load_mgh(PETSurface.mgh)
The problem appears when I compare the values of PET in some vertices. The most of the vertices checked (not all) doesn’t have the same value in tksurfer and matlab (obtained with vol(numberOfVertex)). The values in tksurfer and matlab are similar (in the same magnitude order) but are not equal. Maybe I’m taking a wrong way of obtain the PET activity in Matlab.
I hope that someone can help me. Thank you all in advance, Rafael.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer