Sorry, my last message did not include the previous correspondence. It's pasted below.
Thanks for your help.
Hi,
This is a different large-image bug, that's fixed in our
internal builds - Lars: I'll send you a link to a centos5_64 fslhd
that should allow your analysis to complete..
Many Regards
Matthew
> Hi,
>
> This is likely either a big-RAM problem (in which case see the FSL
> FAQ) or a previous large-image FSL bug. I think the very latest
> version of FSL (4.1.2) may have fixed the latter (Matthew can
> comment further on this maybe) so you may be ok if you just install
> the latest version and try again?
>
> Cheers.
>
>
>
> On 14 Jan 2009, at 08:00, Lars Tjelta Westlye wrote:
>
>> Hi,
>>
>> I'm running TBSS (FSL 4.1.1-centos5_64) on a dataset comprising 433
>> FA
>> volumes. Everything runs fine up to tbss_3_postreg -S subsequent to
>> the
>> fslmerge command at line 164 in the tbss_3_postreg script where it
>> errors
>> attempting to compute the mask and mean FA. That is, all_FA.nii.gz is
>> generated (2.5 GB), but FSL can't seem to read the file:
>>
>> ****
>> [larstw@psyk077 stats]$ fslhd all_FA.nii.gz
>> ** ERROR: nifti_image_open(all_FA): bad header info
>> Error: failed to open file all_FA
>> ERROR: Could not open file
>> ****
>>
>> The same error appears when trying to open the 4D volume in Fslview.
>> However, Freesurfer's mri_info yields the following info, which
>> seems fine
>> to me:
>>
>> ****
>> [larstw@psyk077 stats]$ mri_info all_FA.nii.gz
>> Volume information for all_FA.nii.gz
>> type: nii
>> dimensions: 182 x 218 x 182 x 433
>> voxel sizes: 1.0000, 1.0000, 1.0000
>> type: FLOAT (3)
>> fov: 182.000
>> dof: 0
>> xstart: -91.0, xend: 91.0
>> ystart: -109.0, yend: 109.0
>> zstart: -91.0, zend: 91.0
>> TR: 1000.00 msec, TE: 0.00 msec, TI: 0.00 msec, flip angle:
>> 0.00 degrees
>> nframes: 433
>> ras xform present
>> xform info: x_r = -1.0000, y_r = 0.0000, z_r = 0.0000, c_r =
>> -1.0000
>> : x_a = 0.0000, y_a = 1.0000, z_a = 0.0000, c_a =
>> -17.0000
>> : x_s = 0.0000, y_s = 0.0000, z_s = 1.0000, c_s =
>> 19.0000
>> Orientation : LAS
>> Primary Slice Direction: axial
>>
>> voxel to ras transform:
>> -1.0000 0.0000 0.0000 90.0000
>> 0.0000 1.0000 0.0000 -126.0000
>> 0.0000 0.0000 1.0000 -72.0000
>> 0.0000 0.0000 0.0000 1.0000
>>
>> voxel-to-ras determinant -1
>>
>> ras to voxel transform:
>> -1.0000 0.0000 0.0000 90.0000
>> -0.0000 1.0000 -0.0000 126.0000
>> -0.0000 -0.0000 1.0000 72.0000
>> 0.0000 0.0000 0.0000 1.0000
>> ****
>>
>> I have approx 1TB available disk space. This is the output of free -
>> m (16
>> GB RAM, 18 GB swap):
>>
>> ***
>> [larstw@psyk077 stats]$ free -m
>> total used free shared buffers
>> cached
>> Mem: 16053 4316 11736 0
>> 475 1748
>> -/+ buffers/cache: 2092 13960
>> Swap: 18433 4781 13652
>> ***
>>
>> ***
>> [larstw@psyk077 stats]$ uname -a
>> Linux psyk077.uio.no 2.6.18-92.1.22.el5 #1 SMP Fri Dec 5 09:28:22
>> EST 2008
>> x86_64 GNU/Linux
>> ***
>>
>> Is this expected behavior on datasets of this size? Could the
>> problem be
>> due to insufficient memory on my system, and do you have any
>> suggestions
>> on how to solve this? It should be noted that everything works
>> nicely with
>> smaller datasets, although I would like to include everything in
>> one run.
>>
>> Thank you very much,
>> Lars
>>
>
>
> ---------------------------------------------------------------------------
> Stephen M. Smith, Professor of Biomedical Engineering
> Associate Director, Oxford University FMRIB Centre
>
> FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
> +44 (0) 1865 222726 (fax 222717)
> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve
> ---------------------------------------------------------------------------
|