What is h.all.suvr.sm05.proj0.5.nii.gz? Where did it come from? The name looks like it might be a voxel-wise map, but you say something about ROIs. mri_glmfit does not correct for multiple comparisons. For ROI-based analyses, you can use bonferroni or FDR. For maps you can use mri_glmfit-sim.
On 4/4/2023 4:53 PM, John Anderson wrote:
External Email - Use Caution
Hello Freesurfer, Currently, I am using the Petsurfer manual, which is easily understandable and clear. Nonetheless, I have a question regarding the "mri_glmfit" step. As far as I know, this command is used to correct data for multiple comparisons among all the ROIs in the aparc+aseg.mgz file. Can you please confirm if my understanding is correct?
mri_glmfit.bin --y lh.all.suvr.sm05.proj0.5.nii.gz --fsgd 2G0C.fsgd --C patients_vs_controls.mtx --surf fsaverage lh --cortex --glmdir lh.patinets_vs_controls
In particular, I am interested in correcting data across lobes instead of across all ROIs. I followed the steps mentioned in the FS forum to create lobes, and I can observe lobe annotations in the wmparc.mgz file. I am curious to know whether using the "--annot" flag will adjust the p-values across different lobe ROIs instead of all ROIs in the aparc+aseg.mgz file. I have attempted this, but it appears to only change the name of the annotation in the final statistics file, and the p-value remains the same. Do I need to include any additional arguments in the command?
Thanks for your help John
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer