[Mrtrix-discussion] Track weighted imaging

[Robert Smith]

Hi Aaron (and all),

Sorry for the delayed reply; I had expected Fernando to take this one
himself, but he's been bogged down with other duties.

The tckmap command in the new MRtrix handles the whole generalised concept
of track-weighted imaging (TWI). Track-weighted functional connectivity
(TW-FC) is one particular application within this framework. The relevant
options within tckmap that you need to be aware of to duplicate the method
as implemented in the TW-FC manuscript are as follows:

   - -contrast: This specifies the type of image contrast that is assigned
   to each streamline. In the case of TW-FC, this contrast is drawn from an
   external image source; therefore you want to use scalar_map as the value
   for this option.
   You must then also provide the path to this external image; this is done
   using the -image option.
   Note that this image does not necessarily have to be re-sampled to the
   same voxel grid as the diffusion images; it is adequate for the images to
   simply be aligned in real space, i.e. an anatomical location in the
   diffusion image maps to the same anatomical location in the associated
   scalar image, even if the image transforms / voxel sizes differ.
   - -stat_vox: This controls how the contributions to a voxel from
   multiple streamlines are combined into a single scalar value. In the TW-FC
   manuscript we used the *mean* value across all streamlines traversing
   each grid element.
   - -stat_tck: In cases where the selected form of image contrast may vary
   along the length of each streamline (which is the case for the
scalar_mapcontrast), this option controls how the values computed
along the length of
   a streamline are combined to produce a single scalar value for that
   streamline. In the TW-FC manuscript, the *sum* of these values was used.

The manuscript also described a method for masking out regions from the
TW-FC map with unreliable values due to having too few streamlines traverse
them. This can be achieved using something like:

tckmap streamlines.tck - -template tw_fc.mif -contrast scalar_map_count
-image tw_fc.mif | mrthreshold - - -abs 5 | mrcalc tw_fc.mif - -mult
tw_fc_masked.mif

In case you're not familiar with piping in MRtrix, the standalone '-'
character indicates a temporary image that is to be passed from the output
of one command to the input of another command, and the '|' character
separates the commands.
So in this case, there are three commands called in succession:

   1. tckmap: Generate a track-weighted image where the contrast is simply
   a streamline count (i.e. TDI), but only those streamlines with a non-zero
   FC value contribute to the TDI;
   2. mrthreshold: Generate a binary mask image of those voxels traversed
   by at least 5 streamlines with non-zero FC value;
   3. mrcalc: Multiply the previously-calculated TW-FC map by this mask.

Have fun!
Rob


--

*Robert Smith, Ph.D*
Research Officer, Imaging Division

The Florey Institute of Neuroscience and Mental Health
Melbourne Brain Centre - Austin Campus
245 Burgundy Street
Heidelberg Vic 3084
Ph: +61 3 9035 7128
Fax: +61 3 9035 7301
www.florey.edu.au


On Wed, May 28, 2014 at 6:41 AM, arswitze at ucalgary.ca
<arswitze at ucalgary.ca>wrote:

> Hello Romain,
>
>    Thank you for your reply. Which option within in tckmap would I use to
> create a density volume that weights the tracks by a functional
> connectivity value? In other words, is there an option that could
> combine the tractogram with a functional connectivity map (transformed
> into diffusion space) where the sum of the functional connectivity
> values along a track would be calculated, and that sum would be
> assigned to the entire track. The final map would have a track weighted
> image that is weighted by the associated functional connectivity
> values. This method was outlined in this paper:
> http://www.sciencedirect.com/science/article/pii/S1053811912012402 but
> I have been unable to recreate it.
>
> Thank you for your help,
> Aaron
>
> > Le 27/05/2014 20:01, arswitze at ucalgary.ca a écrit :
> >> Hello mrtrix community,
> >>
> >>      I am fairly new at mrtrix, and diffusion imaging in general, and I
> >> just had a question about the a command that is in mrtrix3. I am
> >> wondering if there is a way to perform track weighted imaging (TWI)
> >> using the newest build of mrtrix3. I was unable to find the command on
> >> the wiki, but in a previous discussion topic regarding the release of
> >> mrtrix3, it was mentioned that this version would include TWI
> >> functionality.
> >>
> >> Thank you for your help,
> >> Aaron Switzer
> >>
> >>
> >> _______________________________________________
> >> Mrtrix-discussion mailing list
> >> Mrtrix-discussion at www.nitrc.org
> >> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
> >>
> > Hello
> >
> > this is the function tckmap that will transform your track into density
> > volume
> >
> >
> > _______________________________________________
> > Mrtrix-discussion mailing list
> > Mrtrix-discussion at www.nitrc.org
> > http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
> >
> >
> >
>
>
> _______________________________________________
> Mrtrix-discussion mailing list
> Mrtrix-discussion at www.nitrc.org
> http://www.nitrc.org/mailman/listinfo/mrtrix-discussion
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.nitrc.org/pipermail/mrtrix-discussion/attachments/20140528/0f9afd8b/attachment.html>