I've added a voxel-based analysis to the wiki page https://surfer.nmr.mgh.harvard.edu/fswiki/PetSurfer

Note that I'm recommending using permutation instead of GRF for correction of multiple comparisons

On 08/29/2017 10:03 AM, John Anderson wrote:

Hi Dr Greve,

I am using pet surfer to analyze suv images. I followed all the steps in wiki, I used the option mgx then I fed the imagesmgx.{ctxgm,subctxgm,gm} into surface based analysis for left and right hemispheres. I want to inquire about the sub cortical analysis.

1. I assume the analysis for sub cortical gray matter (i.e. using the

imagessubctxgm) is volumetric. Is this correct?

Yes. This is on the wiki page.

2. Are the following steps correct for sub cortical gray matter analysis

mri_vol2vol --movsubctxgm.nii.gz --reg reg.dat --tal --talres 2 --talxfm talairach.xfm --nearest --no-save-reg --osubctxgm.tal2mm.nii.gz

mri_masksubctxgm.tal2mm.nii.gz $FREESURFER_HOME/subjects/fsaverage/mri.2mm/subcort.mask.mgzsubctxgm.tal2mm_subc.nii.gz &

Concatenate all subjects together using mri_concat

smooth using mri_fwhm (what is the recommended smoothing for suv)?

Then Group analysis:

mri_glmfit --y all.suvr.tal2mm.subc.sm10.nii --fsgd fsgd.dat --C contrast.mtx --glmdir dir

mri_glmfit-sim --glmdir dir --grf 1.3 pos --cwpvalthresh 0.0166

for "mri_glmfit-sim" in subcortical structure is the flag grf correct or cache?

Thanks in advance! John

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer