Hi FSLers,
I saw this post while searching - I'm wondering exactly how to do this... I
have a run TBSS on a number of controls and patients and have looked at
group differences using a t-test, following the example in the manual. Now,
if I'd like to use ICA, would I just use the mean FA skeleton that I've
already generated? So, I'd just be using the all_FA_skeletonized file.
This doesn't sound right to me, since I don't have a good idea of what ICA
would actually extract from this...
Alternatively, would a better way be to concatenate the individual (and
aligned) FA maps in the FAi directory, then run MELODIC on the concatenated
4D file? This way, MELODIC would extract regions corresponding to the time
series, which would be the individual subjects...
Any thoughts would be appreciated greatly.
Thanks,
Pat
|