Dear FSL users & experts,
This seems like a basic question, so many apologies, but I'm having some trouble with making this work despite numerous attempts and combing the forum's previous questions.
I have used fast to segment subject T1 scans to 3 components, including CSF. I want to use this CSF component as a mask to define a time series in the subject's resting state scan, and then regress it out. However, I'm having trouble transforming it to standard space at 4mm resolution. The key problem is getting the image from native to standard space (once that's done, I think I can just use usesqform to get it to the desired resolution).
I have tried -applyxfm -init /fsl/bin/flirtsch/ident.mat but, this gives me a very displaced image.
I have also tried using -applyxfm usesqform instead of ident.mat, but this gives me an image that is not registered to standard space (though it is not displaced).
(The input images I have used in both cases is the binarised CSF mask I segmented out, and the -ref image is a standard MNI brain transformed to 4mm space using -applyxfm usesqform.)
I would be really grateful if anyone can offer any thought as to why my transformation is not working/ what I need to do instead!
Many thanks in advance for your help.
Kind regards,
Lucia
|