Dear Experts,
We are trying to run vertex wise analysis in Freesurfer to compare two groups.
We are only interested in the prefrontal cortex. Since we need to apply a fdr correction (for multiple testing), we hypothesized that restricting our analyses to the prefrontal cortex will change the level of significance (since we are only analyzing the vertex within the prefrontal cortex instead of the whole brain).
However, when we ran our analysis, it doesn’t seem to be the case.
Here are the different steps that we follow from the Freesurfer website :
https://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/GroupAnalysis
mris_preproc --fsgd $SUBJECTS_DIR/matrices/matrix_QC.fsgd --target fsaverage --hemi rh --meas thickness --out rh.AGS_QC.thickness.00.mgh
mri_surf2surf --hemi rh --s fsaverage --sval rh.AGS_QC.thickness.00.mgh --fwhm 10 --cortex --tval rh.AGS_QC.thickness.10.mgh
mri_glmfit --y rh.AGS_QC.thickness.10.mgh --fsgd $SUBJECTS_DIR/matrices/matrix_QC.fsgd dods --C $SUBJECTS_DIR/contrast/contrast.mtx --surf fsaverage rh --label rh.prefrontal.label --glmdir rh.SZ_HC_thickness.glmdir
mri_fdr --fdr .05 --i /path/SUBJECTS/rh.SZ_HC_thickness.glmdir
/contrast/sig.mgh /path/SUBJECTSrh.SZ_HC_thickness.glmdir
/mask.mgh nooutput
Do you agree that restricting our analysis to the prefrontal cortex should impact (and increase) the level of significance ?
Do you think that these steps are correct ?
Many thanks in advance for your help