Print

---

Hi

Sometimes you can find streamlines meandering through the brain before hitting your inclusion criteria. Because their trajectory is 3D but you only look at them using 2D slices, it is sometimes difficult to appreciate that they actually satisfy your criteria, despite looking like obvious false positives.

One way to check this is to save the paths (using the hidden --save_paths option). You can save only the subset of paths that you are worried about (with the use of an extra waypoint mask for example, which shouldn’t mess up with the randomness of the algorithm) and check in matlab the shape of their trajectory and whether they indeed satisfy your criteria etc.

Cheers

Saad

Dear FSL experts,

I have run probtrackx2 with a seed mask (left SPC) and a target mask (left DLPFC) but I'm in doubt about the correctness of my results.

I'm interested in the proportion of streamlines from the seed that reached the target, as well as in extracting the tract-specific FA values. I used the same mask simultaneously as a target, waypoint and a stop mask. To my understanding, the waypoint mask should restrict the streamlines to only those passing through the mask (in my case the target mask), shouldn't it? However, when I view fdt_paths.nii in fslview, this doesn't seem to be the case (there are some streamlines that leave the seed but don't seem to pass through the waypoint mask). I'd be happy to send you a screenshot from fslview and the output from FDT, if needed.

I'm using FSL Version 5.0.4. and the following command line:
probtrackx2 -x leftSPC.nii -l --onewaycondition --omatrix2 --target2=leftDLPFC.nii -c 0.2 -S 2000 --steplength=0.5 -P 5000 --fibthresh=0.01 --distthresh=0.0 --sampvox=0.0 --xfm=struct2diff.mat --stop=leftDLPFC.nii --forcedir --opd -s merged -m nodif_brain_mask --dir=seed2target --waypoints=waypoints.txt --waycond=AND --targetmasks=targets.txt --os2t

Any help is greatly appreciated!

Best regards,
Lora