Hello,
I am trying to use the following command to analyse crossing fibers in TBSS.
tbss_x F1 F2 D1 D2
It generates all_F(i)_x_skeletonised.nii.gz but not all_F(i)_x.nii.gz (whole brain 4D file...similar to all_FA.nii.gz).
As it has been suggested in S. Jbabdi, T. E. Behrens, and S. M. Smith. Crossing fibres in tract-based spatial statistics. NeuroImage, 49:249-256, 2010, skeletonisation is based on projecting the highest neighboring FA on to the skeleton and FA decreases due to crossing fibres might bias such an analyses. Hence, i was wondering if it makes more sense to perform voxelwise analyses on the whole brain data ( all_F(i)_x.nii.gz ) as opposed to skeletonised data (all_F(i)_x_skeletonised.nii.gz)?
Is it also possible to calculate the percentage of crossing fibers identified by bedpostx? Is it that any value above 0 in all_F2_x_skeletonised.nii.gz is an indication of crossing fiber or do i need to use another threshold?
Best Wishes
Jay
|