Print

Print


right - what you want to do is use your target mask as a "classification target". 
This way, you will have as an output a file called seed_to_<target>, where <target> is one (or several) classification target mask(s). 
These files contain the information you  are looking for, which is "how many sample tracts made it to the target". This info will be voxel-wise, i.e. you will have this number for each seed voxel. You can simply add them using fslmaths/fslstats to get the total number of tracts reaching the target mask.

Cheers,
Saad.


On 2 Jul 2008, at 12:38, Junichi Chikazoe wrote:

Hi Saad,

Thank you very much for your swift service!!

I’m trying to count the number of tracts that cross a particular ROI. The 
most straightforward way to achieve this would be to set a ROI and count 
the tracts that reached the ROI.

However, the default usage of FSL outputs the total number of counts 
across all the voxels within the ROI, which does not necessarily reflect 
the accurate number of the anatomical connections when one tract crosses 
multiple voxels in the ROI.

It would be ideal if pathways stop as soon as they enter a termination 
mask, instead of as soon as they leave a termination mask. I imagine the 
algorithms of the two are quite similar and thus relatively easy to modify 
but I’m not good at programming in linux environment….

If you make your sphere a one voxel thick shell, it would still count 
multiply when a tract proceeds along the margin of the sphere, which I 
have often seen. Also, the use of an exclusion mask might help, but the 
mask does not produce a proper tractography to present.

I would appreciate your thoughts.

Sincerely yours,

Junichi


---------------------------------------------------------------------------
Saad Jbabdi, 
Postdoctoral Research Assistant,  
Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222545  (fax 222717)
[log in to unmask]    http://www.fmrib.ox.ac.uk/~saad
---------------------------------------------------------------------------