Thanks for your answer.
I run FIRST without left accumbens. I obtained for all my structures .bvars, .vtk. For the putamen, the thalamus and the pallidum, I have .nii.gz and the file corrected.  But, for the others structures, I have not the .nii.gz and i have a problem with first boundary correction with the option fast.
This is what I can read on my shell:

** ERROR (nifti_image_read): failed to find header file for 'output_460108-L_Amyg_corr'
** ERROR: nifti_image_open(output_460108-L_Amyg_corr): bad header info
Error: failed to open file output_460108-L_Amyg_corr
Cannot open volume output_460108-L_Amyg_corr for reading!
** ERROR (nifti_image_read): failed to find header file for 'output_460108_all_fast_firstseg'
** ERROR: nifti_image_open(output_460108_all_fast_firstseg): bad header info
Error: failed to open file output_460108_all_fast_firstseg
ERROR: Could not open image output_460108_all_fast_firstseg
Image Exception : #22 :: Failed to read volume output_460108_all_fast_firstseg
Aborted

I don't understand. I know my data are floating data.

For the volume I mentionned (5036 mm3), it was for the left hippocampus, I used the command: fslstats output_460108 -l 16.5 -u 17.5 - V
I had use the boundary correction with the option "none". Perhaps it's for that!

Thanks
Charlotte


> Date: Tue, 4 Jan 2011 11:39:11 +0000
> From: [log in to unmask]
> Subject: Re: [FSL] FIRST error
> To: [log in to unmask]
>
> Hi,
>
> Sometimes FIRST has problems estimating small structures, or structures
> on the edge of the FOV. I suspect that if you only see errors relating to the
> Left Accumbens, then it might be a problem for only this structure. Can you
> test this by running FIRST (via run_first_all - which is what I am assuming
> that you are doing) and specify the option:
> -s L_Amyg,L_Caud,L_Hipp,L_Late,L_Pall,L_Puta,L_Thal,R_Amyg,R_Caud,R_Hipp,R_Late,R_Pall,R_Puta,R_Thal,BrStem
>
> If this works then we can try and sort out the accumbens problem separately.
>
> As for the volume question - what image did you run this on, what command
> did you use, and what structure were you trying to measure? Without this
> information I really cannot comment further.
>
> All the best,
> Mark
>
>
> On 3 Jan 2011, at 16:35, SUBSCRIBE FSL Anonymous wrote:
>
> > Dear FSL users,
> >
> > I try to understand how to run FIRST. I have two problems:
> >
> > 1) I have run it on one of my subject to better understand how the method works.
> > My problem is that I don't have the file: output_name_all_fast_firstseg.nii.gz and I have this error in the logs:
> > ** ERROR (nifti_image_read): failed to find header file for 'output_460108-L_Accu_corr'
> > ** ERROR: nifti_image_open(output_460108-L_Accu_corr): bad header info
> > Error: failed to open file output_460108-L_Accu_corr
> > Cannot open volume output_460108-L_Accu_corr for reading!
> > ** ERROR (nifti_image_read): failed to find header file for 'output_460108_all_fast_firstseg'
> > ** ERROR: nifti_image_open(output_460108_all_fast_firstseg): bad header info
> > Error: failed to open file output_460108_all_fast_firstseg
> > ERROR: Could not open image output_460108_all_fast_firstseg
> > Image Exception : #22 :: Failed to read volume output_460108_all_fast_firstseg
> > Aborted
> >
> > When I wrote cat *.logs/*.e*, i can read that:
> >
> > /usr/local/fsl/bin/first_boundary_corr: line 149: printf: 131.0000000000: nombre non valable
> > /usr/local/fsl/bin/first_boundary_corr: line 149: printf: 131.0000000000: nombre non valable
> > /usr/local/fsl/bin/first_boundary_corr: line 149: printf: 73.0000000000: nombre non valable
> > terminate called after throwing an instance of 'NEWMAT::SingularException'
> > /usr/local/fsl/bin/first_boundary_corr: line 88: 9426 Abandon flirt -in ${imout}_roi -ref ${imlb} -out ${imout} -interp
> >
> > I don't understand what I mean. I believe there is a problem with the boundary correction but I don't know how to correct it.
> >
> > 2) My second problem is with fslstat:
> > I run manually each stage of FIRST with the same subject. I had no problem! But I obtained avolume of 5036 mm3. It's big isn't it?
> >
> > I hope somebody can help me,
> > Thanks
> > Charlotte
> >