Hi again,
you don't specify which version of FAST you were "playing with". If this is the FAST3 version, I would definitely recommend first switching to FAST4 (with the default options), which greatly improves the segmentation of the subcortical structures. However, GPe and GPi are amongst the structures that are the less likely to be recognised as GM, so you might then need to use priors...
Hope this helps,
Gwenaelle
--- En date de : Jeu 3.7.08, Anastasia Ford <[log in to unmask]> a écrit :
> De: Anastasia Ford <[log in to unmask]>
> Objet: [FSL] FAST with apriori probability maps
> À: [log in to unmask]
> Date: Jeudi 3 Juillet 2008, 16h38
> Hello,
>
> I run FAST with apriori probability maps for initialization
> and posteriors.
> However, it seems that the segmentation fails for some of
> the regions of the
> basal ganglia. The thalamus, GPe/GPi appear to be
> classified as white
> matter. Are there any guidelines for improving the
> segmentation?
>
> Thank you,
> Anastasia
_____________________________________________________________________________
Envoyez avec Yahoo! Mail. Une boite mail plus intelligente http://mail.yahoo.fr
