How to use SUIT for VBM
Voxel-based morphometry (VBM) involves a voxel-wise comparison of the local concentration of gray matter between two groups of subjects. This technique is exceptionally useful in the analysis of degenerated or damaged cerebella, one can derive a quantifiable measure of the degeneration/damage when comparing a healthy and unhealthy group. VBM is based on the segmentation map that indicates the probability of a voxel being gray matter.
- This is an old version of the procedure, we are currently working in a new pipeline!
- Make sure the anterior commissure is set to 0,0,0.
- Isolate the cerebellum to generate the segmentation
map using suit_isolate('<name>.nii')
. This will also provide you with a gray-matter segmentation map called
<name>_seg1.nii`. - Hand correct the isolated map if necessary, excluding any gray matter included outside the cerebellum.
- Normalize the cropped anatomical images, using
suit_normalize
orsuit_normalize_dartel
. - Reslice the segmentation map into atlas space. For
this purpose, the option 'preserve' has to be set to one, such
that the overall probability mass will be the same after normalization. Use:
suit_reslice ('<name>_seg1.nii', 'm<name>_snc.mat', 'mask', '<name>_pcereb_corr.nii', 'preserve', 1, 'vox', [1 1 1]);
Latter can also be submitted over the SPM job manager:
job.subj.affineTr={['Affine_<name>' subjName '_seg1.mat']}; % Affine transformation
job.subj.flowfield={['u_a_<name>' subjName '_seg1.nii']}; % This is the flowfield
job.subj.resample{1}={['<name>' subjName '_seg1.nii']}; % Gray matter image.
job.subj(s).mask={['c_struct<name>' subjName '_pcereb_corr.nii']}; % Mask
job.jactransf=1; % Imporant for VBM
suit_reslice_dartel(job);
Doing VBM with SUIT has two main advantages: First, the overlap of cerebellar structures is improved, and second, by masking the image before reslicing it into the atlas space, no supra-tentorial gray matter can bias the results.
More information on VBM can be found here.