[Mrtrix-discussion] streamtrack segmentation fault

[Donald Tournier]

Hi Peter,

Not sure I completely get what is going on. When you say "this is a
resampled image", do you mean that the CSD4 file has been regridded in some
way? Or the original DWI image from which the CSD4 file was generated? And
was the regridding done for upsampling the data or to reorient the images?
In any case, regardless of what processing was actually done, there is the
possibility that your seed.mif image no longer maps onto the CSD4.mif image
correctly, so that some of the seed points end up outside the volume of the
CSD4.mif image - I have a feeling streamtrack doesn't check for this, which
would indeed cause a segfault. Bear in mind that streamtrack does its
processing in scanner coordinates, so having a seed image that is 'correct'
voxel-wise with respect to the CSD image is not sufficient if the image
transform matrices don't also match.

Can you post the output of "mrinfo CSD4.mif" and "mrinfo seed.mif"? I just
want to check whether the transform matrices are the same. Also, try to
load CSD4.mif in mrview, and overlay seed.mif using either the ROI or
overlay sidebar tools, to check for alignment.

Hopefully that'll be the problem. If so, I'll add an explicit bounds check
in the code to make sure this doesn't happen for the next release.

Cheers,

Donald.


On 17 October 2012 16:02, Peter Neher <p.neher at dkfz-heidelberg.de> wrote:

>  Hi everyone,
>
> I am trying to use streamtrack (probabilisatic with CSD) on one of my
> datasets and get the following error:
>
> streamtrack [DEBUG]: reading key/value file "/etc/mrtrix.conf"...
> streamtrack [INFO]: opening image "CSD4.mif"...
> streamtrack [DEBUG]: reading key/value file "./CSD4.mif"...
> streamtrack [DEBUG]: sanitising transformation matrix...
> streamtrack [DEBUG]: preparing file "./CSD4.mif"
> streamtrack [DEBUG]: setting up image "CSD4.mif"...
> streamtrack [DEBUG]: sanitising transformation matrix...
> streamtrack [DEBUG]: setting up data increments for "CSD4.mif"...
> streamtrack [DEBUG]: data increments initialised with start = 0, stride =
> [ 15 1395 161820 1 ]
> streamtrack [INFO]: error parsing spherical ROI specification "seed.mif" -
> assuming mask image
> streamtrack [DEBUG]: mapping image "CSD4.mif"...
> streamtrack [DEBUG]: file "./CSD4.mif" mapped at 0x7fa59c174000, size
> 29775147 (read-only)
> streamtrack [DEBUG]: data mapper for image "CSD4.mif" mapped with segment
> size = 7443720 (optimised)
> streamtrack [INFO]: launching 12 threads
> streamtrack [INFO]: opening image "seed.mif"...
> streamtrack [DEBUG]: reading key/value file "./seed.mif"...
> streamtrack [DEBUG]: sanitising transformation matrix...
> streamtrack [DEBUG]: preparing file "./seed.mif"
> streamtrack [DEBUG]: setting up image "seed.mif"...
> streamtrack [DEBUG]: sanitising transformation matrix...
> streamtrack [DEBUG]: setting up data increments for "seed.mif"...
> streamtrack [DEBUG]: data increments initialised with start = 92, stride =
> [ -1 93 10788 ]
> streamtrack [DEBUG]: mapping image "seed.mif"...
> streamtrack [DEBUG]: file "./seed.mif" mapped at 0x7fa5a12b1000, size
> 496501 (read-only)
> streamtrack [DEBUG]: data mapper for image "seed.mif" mapped with segment
> size = 496248
> ./invivo_start.sh: line 14: 12913 Segmentation fault      (core dumped)
> streamtrack SD_PROB CSD4.mif fibers.tck -seed seed.mif -debug &>log.txt
>
> streamtrack is running fine on the original dataset. This is a resampled
> image. But the CSD seems to run just finy, only the tracking fails.
> I am using Ubuntu 12.4 and mrtrix-0.2.10_2012-02-10.
> Any suggestions?
>
> Best,
> Peter
>
> _______________________________________________
> Mrtrix-discussion mailing list
> Mrtrix-discussion at www.nitrc.org
> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>
>


-- 
*Dr Jacques-Donald Tournier
*
Research Fellow

The Florey Institute of Neuroscience and Mental Health
Melbourne Brain Centre - Austin Campus
245 Burgundy Street
Heidelberg  Vic  3084
Ph:  +61 3 9035 7033
Fax:  +61 3 9035 7307
www.florey.edu.au
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.nitrc.org/pipermail/mrtrix-discussion/attachments/20121017/30e9dbda/attachment.html