[Mrtrix-discussion] tracking with high resolution seed/mask/include

[Donald Tournier]

Hi Maarten,

You're right, streamtrack should not have any issues using ROIs with
different resolutions etc. In my experience, the only time this kind of
'freeze' happens is when no tracks are being accepted. This can mean one of
two things:

(1) none of the seed locations are acceptable (every point/direction sampled
from the seed ROI has FOD amplitude lower than the threshold). This is
unlikely to be the case for you given that you seem to be seeding from the
whole white matter segment. By the way, does the tracking work OK if you use
the 1mm WM seed and brain mask ROIs and the 2mm target ROI? The other thing
to check is that the amplitude of the 1mm ROIs is greater than 0.5 in the
appropriate regions (this is the threshold used to decide what is and isn't
in the ROI). How about if you use the 1mm ROIs but don't supply a target?
Presumably this produce whole-brain tracks? If not, it would point at a
problem with the seed or mask ROIs.

(2) none of the tracks generated satisfy the inclusion criteria. Either they
are all shorter than the default 10mm minimum distance threshold (unlikely),
or none of them make it to the target ROI. My guess is the latter. Can you
check whether the tracks produced using the 2mm ROIs go right through the
centre of the target ROI, or whether they just skim the edges? Maybe the
extent of the 2mm ROI is larger by just the right amount for some of the
tracks to enter it. One thing you could try is to dilate the target ROIs -
you can use the command "erode -dilate in out" to do this (I agree the name
is a little unfortunate here...).

Other than that, I can't see where the problem might be...
Cheers,

Donald.


On 9 April 2011 00:37, Maarten Vaessen <m.vaessen at gmail.com> wrote:

> Hello,
>
> i am trying to do a streamtrack with SD_PROB using high resolution seed
> mask and include images. The DWI data is at 2mm iso, the other masks are at
> 1mm iso from the T1 image, they are from a freesurfer segmentation. I
> checked whether all files are in good registration with mrview (and they
> are).
>
> i run something like: streamtrack SD_PROB CSD_2mm.nii -seed wm_1mm -mask
> brain_1mm -include target_1mm -num 10000 tracks.tck
>
> Alas, the program freezes after execution, there are no errors or messages.
> Just no output and nothing happens. The file tracks.tck is created on the
> hd, although its very small in size (420 bytes).
>
> When i try the same with 2mm mask in DWI space, all runs fine.
>
> I checked the source code to see what could go wrong, all i could find is
> that base.cpp makes Mask objects which are in principle Interp objects, so
> getting a value from the Mask from a tracking point in mm coords should be
> no problem right?
>
> kind regards Maarten
>
> _______________________________________________
> Mrtrix-discussion mailing list
> Mrtrix-discussion at www.nitrc.org
> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>
>


-- 
Jacques-Donald Tournier (PhD)
Brain Research Institute, Melbourne, Australia
Tel: +61 (0)3 9035 7033
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.nitrc.org/pipermail/mrtrix-discussion/attachments/20110411/01e52a06/attachment.html