FSL Experts,
I am writing to follow-up on a posting regarding multiple regression in TBSS. Similar to the 2009 posting (https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0903&L=FSL&P=R65741&1=FSL&9=A&J=on&d=No+Match%3BMatch%3BMatches&z=4), I have just run a multiple regression with tbss on skeletonised FA data in a single group of 22 subjects using 4 EVs. My primary EV is a neuropsychological/behavioral variable and I would like to control for age, sex, and education (other 3 EVs). I created a design matrix using the FSL GLM with my variable of interest (EV1) and 3 covariates (EV2-EV4), which have been manually demeaned. I hoped that this would show significant relationships between white matter and my behavioral variable with the 3 covariates controlled for. This was using randomise_parallel and the TFCE option. When I display the tbss_tfce_corrp_tstat contrast image in fslview with the -b 0.95,1 threshold, however, most of the skeletonised data appears to correlate with the target variable. This is the case even when I use a more conservative 0.99,1 threshold.
I highly doubt that the entire brain is correlated with my behavioral variable, so I was hoping to get some guidance on what I might be doing wrong. I have attached snapshots of my GLM setup and contrast tab.
Any guidance you can provide would be GREATLY appreciated.
Thank you so much for your time,
Cristina
|