Dear Dr Doug, I have two groups of subjects (80 patients and 40 controls). I analyzed the data using Freesurfer v6.0. I want to study the difference between the groups in "precentral gyrus volume". After "recon-all" I built Qdec table and I used Qdec to compute the comparison by including eTIV a covariate. After correcting the results using Monte Carlo simulation or adjusting for FDR I found no difference between the groups in volume neither in surface area.
I adjusted the precentral gyrus volume values for eTIV using the commands:
mri_label_volume -eTIV $SUBJECTS_DIR/${i}/mri/transforms/talairach.xfm 1948 $SUBJECTS_DIR/${i}/mri/wmparc.mgz 2017 mri_label_volume -eTIV $SUBJECTS_DIR/${i}/mri/transforms/talairach.xfm 1948 $SUBJECTS_DIR/${i}/mri/wmparc.mgz 2024
I extracted the adjusted volume values to SPSS then I ran box plots comparison between the groups and I found difference at (P<0.0001) Why Qdec doesn't show the difference between the groups although the p value is very small. What I am doing wrong in Qdec? Is it wrong to include eTIV a covariate?
Thank you for any advice John