Hi Antonella,
I am doing a DTI group analysis study more exactly I am interested to see any changes in the FA values within two groups: controls versus patients (children with epilepsy). I introduced a white matter mask that will ensure that I am only considering diffusion values in the white matter and then also I did corrected for multiple comparison running mri_glmfit-sim with the option grf (as Doug recommend me).
My questions are:
1)How can I conclude if there is an increase and/or decrease in the FA values in one group (for example patients) versus another group (example controls). In other words what I can conclude regarding the FA values based on the GLM results: blue versus red clusters from the http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/Diffusion (I don't see anymore the image with red and blue clusters posted on this website!)
See my previous email.
- If my fa-mask was defined using the wmparc2diff.mgz so is a segmentation volume, in order to resample this volume (wmparc or aseg) in CVS space, do I need to use the nearest neighbor (nearest) interpolation option instead of --interp trilin? What will be the
difference and is the difference visible in the intensity or in another parameter such as a smoother registration?
Yes, with any of the segmentation / parcellation file you would want to use the nearest neighbor interpolation option. The difference (if you used trilinear instead) would be visible. You would see new and uninterpretable segmentation lables in the resampled volume which are not valid if you want to interpret them as anatomical references. (The interpolation does not change the registration results though, it just affects the way your resampling gets executed.)
--Lilla