Hi,
I'm afraid this is an intermittent bug within FIRST that occurs only in the brainstem when the FOV is slightly tight at the base of the brainstem. It doesn't always occur, which is why we haven't yet fixed it.
If you are not that interested in the brainstem then you can just leave this out of the list of structures, as the bug does not affect the segmentation of the other structures.
If you do really want the brainstem then sometimes it can be fixed by extending the inferior portion of the image with copies of the most inferior slice (you can do this with fslroi and fslmerge).
All the best,
Mark
On 28 Jun 2013, at 23:07, Ran Xu <[log in to unmask]> wrote:
> Dear FSL experts,
>
> I ran the run_first_all command, then it appeared an error during the segmentation part, like "first crashed with SIGSEGV". I can get other subcortical segmentation except for the BrainStem. When the command executes the segmentation and the mask is BrStem_bin.bmv, the error comes out, but for other parts, it works well.
> I also tried run_first command, it still can't work. The following shows the same segmentation error.
> Could anyone give me some advice? Is this a system problem or something else?
>
> Thanks in advance.
>
> run_first -i con02 -t con02_to_std_sub.mat -n 40 -o con02-Brain -m ${FSLDIR}/data/first/models_336_bin/BrStem_bin.bmv
> create shapeModel
> done creating shapeModel
> 0.892234 -0.0368041 0.0277029
> 0.0612831 0.886331 -0.143025
> -0.017489 0.16302 0.850447
> NEw done imodes transform
> Segmentation fault (core dumped)
>
