Hi Cherif,
there is maybe another way of doing the things, but why don't you intersect your thresholded p-corrected map with the probabilistic tracts from either JHU or Jülich white matter atlases? That would make it an anatomically relevant way of segregating your clusters...
Cheers,
Gwenaëlle
--- En date de : Ven 23.1.09, Cherif Sahyoun <[log in to unmask]> a écrit :
> De: Cherif Sahyoun <[log in to unmask]>
> Objet: [FSL] FA per cluster
> À: [log in to unmask]
> Date: Vendredi 23 Janvier 2009, 20h13
> Hi,
>
> I'm trying to automate getting the mean FA from each
> subject for each output
> contrast of TBSS, within each cluster.
>
> If I binarise my contrast (at a certain threshold) and then
> apply it as a
> mask to all_FA_skeletonised with fslmeants, I get the mean
> FA for all sig.
> clusters together, but I'd like to break it down to a
> per-cluster output.
>
> Using the cluster command finds the clusters, but does not
> output individual
> images for each.
>
> Ideally I'd like a way to use the clusters from
> "cluster" as masks in
> fslmeants, so that I'd get one big table for each
> contrast... any thoughts?
> (or I should say any thoughts that would not involve
> drawing individual ROIs
> by hand :) )
>
> Thanks,
> Cherif
|