Dear FSL Experts,
Recently we are analyzing some DTI data, using the commandline below for randomise(group difference, 2 groups unpaired t-test):
randomise -i all_FA_skeletonised -o tbss -m mean_FA_skeleton_mask -d design.mat -t design.con -n 5000 --T2
We didn't get any significant result, then we changed the --T2 to -T, getting some significant results, corrected:
randomise -i all_FA_skeletonised -o tbss -m mean_FA_skeleton_mask -d design.mat -t design.con -n 5000 -T
From the randomise instruction, we know that --T2 option is optimal for TBSS, why in our case, we acquired more significant outputs from -T than --T2? Is it unusual?
From some old posts, -T option can also be used for TBSS, even though it's not optimal for that:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0802&L=FSL&P=R46028&1=FSL&9=A&I=-3&J=on&K=8&X=7B974CB95260075590&Y=zhangjianeave%40gmail.com&d=No+Match%3BMatch%3BMatches&z=4
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind0708&L=FSL&P=R28295&1=FSL&9=A&I=-3&J=on&K=9&X=7B974CB95260075590&Y=zhangjianeave%40gmail.com&d=No+Match%3BMatch%3BMatches&z=4
Does these mean that we can instead use -T for --T2 for TBSS in some cases?
Best regards,
Jian
|