Hi Mike,
Yep, you won't get any output from cat *e* for a FIRST analysis that
finished successfully (that command is just to cat the error logs).
In the case here, it's failing on the brain stem segmentation. Is your
FOV tighter around the base of the brainstem for these 3 cases? Is
this a structure that you're interested in? If not, the easiest thing
is to just exclude it by running something like
run_first_all -i input -o output -s
L_Accu,L_Amyg,L_Caud,L_Hipp,L_Pall,L_Puta,L_Thal,R_Accu,R_Amyg,R_Caud,R_Hipp,R_Pall,R_Puta,R_Thal
If you are interested in the brain stem, you may be able to "fix" it
by extracting the most inferior slice (using fslroi) and then adding
some copies of it to your original image (using fslmerge).
One caveat regarding the first option, as FIRST ensures that each
voxel is assigned only one value (i.e. a voxel can't be classified as
two separate structures in the final segmentation), you can see slight
differences in the final segmentation based on whether or not you
include/exclude certain structures. In case you are interested, try
the run_first_all command above (excluding the brain stem) on one of
the cases that did run successfully before and see for yourself.
On Tue, Nov 11, 2014 at 8:35 PM, Mike Sugarman <[log in to unmask]> wrote:
> Thank you for your response. As far as I can tell, the registrations look good. I am running FSL 5.0.7; I get the following (long) message when I use the cat *e* command (note that I get no output when attempting this command with files where FIRST ran successfully):
>
> /usr/local/fsl/bin/run_first: line 165: 12837 Segmentation fault: 11 ${FSLDIR}/bin/first -i $image -l $trmat -m $model -k $outname -n $modes $verbose $comExtras
> /Users/AnalyzeData/Desktop/fsl/bin/run_first: line 164: 22649 Segmentation fault: 11 ${FSLDIR}/bin/first -i $image -l $trmat -m $model -k $outname -n $modes $verbose $comExtras
> Cannot open volume -t for reading!
> Cannot open volume -t for reading!
>
> Part of FSL (build 507)
> first_mult_bcorr (Version 1.0) University of Oxford (Mark Jenkinson)
> Converts
>
> Usage:
> first_mult_bcorr [options] -i <T1_image> -c <4D_corrected_labels> -u <4D_uncorrected_labels> -o <output_image>
>
> Compulsory arguments (You MUST set one or more of):
> -i,--in filename of original T1 input image
> -o,--out output image name (3D label image)
> -u,--uncorrected filename of 4D image of uncorrected labels (with boundaries)
> -c,--corrected filename of 4D image of individually corrected labels
>
> Optional arguments (You may optionally specify one or more of):
> -v,--verbose output F-stats to standard out
> -h,--help display this message
>
>
>
> -t: Option doesn't exist!
> Image Exception : #22 :: ERROR: Could not open image FIRST_-BrStem_corr
> terminate called throwing an exception/bin/sh: line 1: 22698 Abort trap: 6 /Users/AnalyzeData/Desktop/fsl/bin/fslmerge -t FIRST__all_fast_firstseg FIRST_-L_Accu_corr FIRST_-L_Amyg_corr FIRST_-L_Caud_corr FIRST_-L_Hipp_corr FIRST_-L_Pall_corr FIRST_-L_Puta_corr FIRST_-L_Thal_corr FIRST_-R_Accu_corr FIRST_-R_Amyg_corr FIRST_-R_Caud_corr FIRST_-R_Hipp_corr FIRST_-R_Pall_corr FIRST_-R_Puta_corr FIRST_-R_Thal_corr FIRST_-BrStem_corr
> Image Exception : #22 :: ERROR: Could not open image FIRST_-BrStem_first
> terminate called throwing an exception/bin/sh: line 1: 22699 Abort trap: 6 /Users/AnalyzeData/Desktop/fsl/bin/fslmerge -t FIRST__all_fast_origsegs FIRST_-L_Accu_first FIRST_-L_Amyg_first FIRST_-L_Caud_first FIRST_-L_Hipp_first FIRST_-L_Pall_first FIRST_-L_Puta_first FIRST_-L_Thal_first FIRST_-R_Accu_first FIRST_-R_Amyg_first FIRST_-R_Caud_first FIRST_-R_Hipp_first FIRST_-R_Pall_first FIRST_-R_Puta_first FIRST_-R_Thal_first FIRST_-BrStem_first
> Image Exception : #22 :: ERROR: Could not open image FIRST__all_fast_origsegs
> Aborted
>
>
--
Niels Bergsland
Integration Director
Buffalo Neuroimaging Analysis Center
100 High St. Buffalo NY 14203
[log in to unmask]
|