Dear FSL experts,
I'm performing 2x2 repeated measures analysis with TBSS. I'd like to test if the detected changes in DTI measures between pre and post treatment are associated with scores at a clinical test. To test that I would use flsmaths in order to subtract the pre-treatment DTI measures images (i.e. dti_FA, dti_L1, dti_L2L3, and dti_MD) of each subject to the post-treatment images, and then use fslmerge to merge together these new subjects' images to perform correlation analyses with randomise, as I would do for fMRI data.
My question is the following: is it correct to do that on DTI structural data? Otherwise, do you have any suggestions?
Thanks for your help,
Elisa
|