Hi - if the idea is to use implant coordinates as seed points, then you do not need to create volume-like masks for that, you can use probtrackx in "simple" mode and input the coordinates as an ascii text file.
For example:
probtrackx -x coords.txt --mode=simple [other options]
where coords.txt looks like this:
x1 y1 z1
x2 y2 z2
...
x,y,z being *voxel* coordinates in seed space (better to use structural space in your case i guess).
If you are interested in estimating the probability of connecting to a set of targets (as in seed-to-target), then you can use the --s2tastext option. The output will be a text file that represents a matrix with size (#seed coords)X(# targets).
Cheers,
Saad.
On 13 Apr 2011, at 12:18, Cinzia Cecchetto wrote:
> Dear FSL experts,
>
> I have to process a set of DTI images of patients. These patients have several implants in the deep brain and I want to know the fibers between these areas (I'm not interested about images of fiber but numbers, density, length ecc), I know everything about these points.
> I thought to use PROBTRACKX of FSL: I completed eddy current correction and bedpostx but I have some problems to insert multiple masks. Do I have to create a 3D images for each ROI? (there are about 100 ROI in each patients) or are there any methods to automatically insert the coordinates? Which is the better space (diffusion, structural or standard space) to create seed masks? Which file format do I have to use to write the masks?
> Bests
>
> Cinzia
>
--
Saad Jbabdi
University of Oxford, FMRIB Centre
JR Hospital, Headington, OX3 9DU, UK
(+44)1865-222466 (fax 717)
www.fmrib.ox.ac.uk/~saad
|