Dear FSL team:
I am having some trouble with FAST in FSL 4.1.5. The segmentation of a
fairly normal-looking 3T SPGR creates a "whole" of CSF inside the frontal
WM. I am attaching some slice screenshots with the problem area marked by a
read circle in the SPGR.
I am running fast as
fast -A prior0.nii prior1.nii prior2.nii -P -v --nopve -B -b -p -o image
I am also attaching the input WM prior and output WM probability image
slices: the prior is virtually 100% certain of WM in the problem area, so
that does not seem to explain the problem. Also, if I run fast with no
priors, I get essentially the same result in that area.
I'd appreciate any suggestions, if this seems to be a problem related to the
way I am running fast.
I would also be happy to upload the original SPGR image and prior maps, if
anyone wants to try this hands-on.
Thanks in advance!