[Mrtrix-discussion] Human Connectome Project dataset issue

[Ivan Alvarez]

Hello Donald,

Thank you for your reply. Regarding your first question, I was feeding 
all 3 shells, I was not aware this was not supported in MRtrix. Is there 
no way of utilising all the information from the 3 shells at present?

As for the second part, you are entirely correct, it seems to be an 
incorrect FOD causing the program to keep looking for a start point. I 
re-computed with FOD with a single shell (b=3000), and streamtrack is 
back to working fine.

Kind regards,
Ivan Alvarez

PhD Candidate
Imaging and Biophysics Unit
UCL Institute of Child Health
30 Guilford Street, London, WC1N 1EH

On 13/01/13 22:57, Donald Tournier wrote:
> Hi Chris & Ivan,
>
> OK, I think there are a few issues to consider here. The first thing 
> is to figure out why the response function estimated for these data is 
> not right. Are you using just a single shell of the HARDI data, or are 
> you simply feeding all 3 shells to MRtrix? MRtrix is only designed to 
> handle single shell data, so if you're doing the latter, it's no 
> surprise that it will have issues. It may be possible to analyse these 
> kinds of data in a future version, but for now you should stick to 
> single-shell HARDI, unfortunately.
>
> As to why the tracking hangs: the most likely reason is that the 
> program is failing to find a suitable seed point. The procedure in 
> MRtrix is to select a random point and a random orientation, compute 
> the FOD amplitude, and start tracking if that amplitude is greater 
> than the -init_cutoff threshold (by default, 0.2). If your response 
> function is incorrect, there is a good chance that the scaling of the 
> resulting FODs will also be off, giving you smaller FODs than expected 
> - in which case streamtrack will try in vain to find a single suitable 
> seed point/orientation, and appear to hang. To verify this, you could 
> try using a -cutoff value of zero to see if that at least produces 
> tracks. You could also use the find_SH_peaks command on the CSD output 
> file, and check the FOD peak amplitudes (although it's a little more 
> work to figure out how to extract the amplitude from that).
>
> If none of this helps, I'm happy to have a look at the data - usually 
> quicker to troubleshoot that way...
>
> Cheers,
>
> Donald.
>
>
> On 12 January 2013 02:49, Parker, Christopher 
> <christopher.parker.10 at ucl.ac.uk 
> <mailto:christopher.parker.10 at ucl.ac.uk>> wrote:
>
>     Hi Ivan and Mrtrix users,
>
>     I have just experienced the same problem with streamtrack SD_PROB.
>     It does not produce any streamlines and hangs although still uses
>     a lot of CPU. I think it might have something to do with the SH
>     harmonic coefficient file (source argument in streamtrack), since
>     when I use a harmonic coefficient file from a different scan it
>     works fine (all other arguments were the same files). I need to
>     find a solution to this also.
>
>     Best,
>     Chris
>     ------------------------------------------------------------------------
>     *From:* mrtrix-discussion-bounces at www.nitrc.org
>     <mailto:mrtrix-discussion-bounces at www.nitrc.org>
>     [mrtrix-discussion-bounces at www.nitrc.org
>     <mailto:mrtrix-discussion-bounces at www.nitrc.org>] on behalf of
>     Ivan Alvarez [ivan.alvarez.11 at ucl.ac.uk
>     <mailto:ivan.alvarez.11 at ucl.ac.uk>]
>     *Sent:* 09 January 2013 15:03
>     *To:* mrtrix-discussion at www.nitrc.org
>     <mailto:mrtrix-discussion at www.nitrc.org>
>     *Subject:* [Mrtrix-discussion] Human Connectome Project dataset issue
>
>     Hello Mrtrix users,
>
>     I tried analysing a sample subject from the recently available
>     Human Connectome Project dataset (https://db.humanconnectome.org),
>     and encountered an issue with CSD tractography in Mrtrix. In
>     brief, the dataset has 1.25 mm isotropic voxels, 270 directions, 3
>     b-shells (1000,2000,300) and 2 repeats.
>
>     The tensor estimation worked fine, and am able to compute
>     sensible-looking FA maps. The mask erosion is also fine, but when
>     I view the response function coefficient, it displays an unusual
>     shape (RCF <http://postimage.org/image/cl7tpwc3p/>). The csdeconv
>     command returns no errors and the orientation plot looks
>     reasonable. My problem is, when I attempt to perform tractography
>     with the streamtrack command, the program hangs.
>
>     Any idea as to why this might be?
>
>     -- 
>     Kind regards,
>     Ivan Alvarez
>
>     PhD Candidate
>     Imaging and Biophysics Unit
>     UCL Institute of Child Health
>     30 Guilford Street, London, WC1N 1EH
>
>
>     _______________________________________________
>     Mrtrix-discussion mailing list
>     Mrtrix-discussion at www.nitrc.org
>     <mailto: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 <http://www.florey.edu.au>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.nitrc.org/pipermail/mrtrix-discussion/attachments/20130114/c2f49c92/attachment.html