Hi everyone,
I ended up creating masks by going into FSL view, selecting a standard space, turning on atlas tools, selecting the relevant structure(s), turning off the visibility of the standard space brain, saving the relevant structure in a file, binarising the file of selected structure (as it looked like a probability map of activation in that region). I used lots of files consisting of binary ROI's like this as my classification targets.
I tried using FEAT query, but it obviously wanted me to create .feat directory, which would make any sense because I was not going to consider fitting a GLM to DTI data. Also it looked like FEAT query was using the same technique, ie the same structures and atlases.
I did use reccomened commands like:
fslmaths binoccipitalLobe7 -kernel sphere 15 -dilD biggerOcciptal2
to inflate ROIs I had created that seemed to small.
Cheers Russell.
|