Dear Dr Greve, thank you so much for answers to my questions.

Regarding question 2, yes I meant leaving only columns I need, for instance I am interested in lobes. I ran these commands on my data

# mri_annotation2label --subject subject --hemi lh --lobesStrict lobes
# mri_annotation2label --subject subject --hemi rh --lobesStrict lobes
# mri_aparc2aseg --s subject --labelwm --hypo-as-wm --rip-unknown \
#   --volmask --o wmparc.lobes.mgz --ctxseg aparc+aseg.mgz \
#   --annot lobes --base-offset 200 [--base-offset must be last arg]

then mri_segstats and I got the final stats file which contain data for the lobes rois

3201    wm-lh-frontal-lobe              
3203    wm-lh-cingulate-lobe       
3204    wm-lh-occiptal-lobe              
3205    wm-lh-temporal-lobe           
3206    wm-lh-parietal-lobe                
3207    wm-lh-insula-lobe                 

4201    wm-rh-frontal-lobe                
4203    wm-rh-cingulate-lobe             
4204    wm-rh-occiptal-lobe              
4205    wm-rh-temporal-lobe             
4206    wm-rh-parietal-lobe                
4207    wm-rh-insula-lobe

My question is can I revise the spreadsheet from the output of aparcstats2table or asegstats2table to keep the lobar ROIs above (8 columns) then feed that data to mri_glmfit?

On 12/20/2022 12:22 PM, James Brown wrote:

Dear Freesurfer experts,
I have 2 questions about the command mri_glmfit and I appreciate your help.
1- Are there any settings or flags that can be used in mri_glmfit to restrict statistical comparison of surface based analysis to specific ROIs rather than whole brain analysis. I would like to show the difference on the brain surface in lobes rather than in sub lobar labels.

You can create a mask or label and feed it into mri_glmfit with --mask or --label

2- I understand that the flag "--table" in command mri_glmfit allows for feeding spreadsheets instead of images. Can we restrict the number of columns in this spreadsheet to 8 columns for lobs (frontal, temporal, occipital and parietal) left and right? Also, is there any way to generate a statistical map from the spreadsheet statistics?

Remove the columns you don't want?

On Tue, Dec 20, 2022 at 12:22 PM James Brown <jb1979000@gmail.com> wrote:

Dear Freesurfer experts,
I have 2 questions about the command mri_glmfit and I appreciate your help.
1- Are there any settings or flags that can be used in mri_glmfit to restrict statistical comparison of surface based analysis to specific ROIs rather than whole brain analysis. I would like to show the difference on the brain surface in lobes rather than in sub lobar labels.
2- I understand that the flag "--table" in command mri_glmfit allows for feeding spreadsheets instead of images. Can we restrict the number of columns in this spreadsheet to 8 columns for lobs (frontal, temporal, occipital and parietal) left and right? Also, is there any way to generate a statistical map from the spreadsheet statistics?

Thank you and happy holidays :)
James