Hi Peggy - I looked at the data. Unfortunately, the voxel size is 0.9x0.9x4 in the first time point and 0.9x0.9x5.2 in the second. So not only is the resolution extremely anisotropic, which is problematic for tractography, but it's also different between scans, which almost guarantees that you will find longitudinal changes (except you won't know if it's because the brain is changing or the scan is changing).
Finally, the 12 directions that you have may not be sufficient to fit a crossing-fiber model like the ball-and-stick model, which may explain the noisy output.
a.y
On Mon, 8 Sep 2014, Peggy Skelly wrote:
Hello Tracula experts, I was wondering if you could comment on the quality of the tracts reconstructed with our 1.5T data. These 2 files are from the same subject at 2 timepoints, processed in the longitudinal stream.
Is much of the non-smooth look of the tracts due to the resolution of the dwi scans (1.8x1.8x4mm)? What could be the source of the diagonal lines/gaps in the tracts seen in tp1 when viewed from above?
Thanks, Peggy