Dear FSL Experts,
I've been running probtrackx2 on HCP data using the 264 ROI's from Power et. al. 2011. When I normalized my values by dividing each row by the waytotal (using FDT_Network_Matrix./repmat(waytotal,1,264) - is this the right way to normalize?) I end up with almost all of my values falling below .3, with relatively sparse connectivity.
1) Are there any suggestions for how to improve the output? Could the ROI's that I am selecting be the problem?
2) What is the range for --samvox? My understanding is that it lets you track from somewhere not in the center of the voxel. How do I set it to jitter the starting location?
probtrackx2 --network -x /mnt/engram/projects/~/data/leviDiff/038/test2/TS.txt -l --omatrix1 -c .2 -S 2000
--steplength=.5 -P 5000 --fibthresh=0.01 --distthresh=0.0 --xfm=/mnt/engram/projects/~/data/leviDiff/disk1/130922/Dif
fusion/data.bedpostX/xfms/standard2diff.mat --invxfm=/mnt/engram/projects/~/data/leviDiff/disk1/130922/Diffusion/data.bedpostX/xf
ms/diff2standard.mat --forcedir --opd -s /mnt/engram/projects/~/data/leviDiff/disk1/130922/Diffusion/data.bedpostX/merged -m /mnt
/engram/projects/~/data/leviDiff/disk1/130922/Diffusion/data.bedpostX/nodif_brain_mask --dir=/mnt/engram/projects/~/data/
leviDiff/disk1/130922/Probtrackx --seedref=/mnt/engram/projects/~/data/leviDiff/avg152T1_brain.nii.gz --sampvox=0
Thank you so much,
LS
|