Tract generation very poor

[mwagshul]

Hi, I'm new to this software, we have been working with TRACULA for a long time (it works well, but have questions about its reproducibility), and wanted to see how this performed in comparison. We have a dataset of about 250 individuals, older adults with and without MS.

Data acquisitions are similar to the ADNI protocol (3T, 2 mm iso, MB3, 16 @ 500, 32 @ 1000, and 60 @ 2000, with full reversed PE scans), data have already run through topup and eddy.

The first two subjects we've run look pretty bad, most of the tracts are very sparse, here's one example of the cingulum.

Here is the call to TractSeg:

TractSeg -i 103_104_NODDI_EddyCorrected.nii.gz --bvals 103_104_NODDI_combined.bvals --bvecs 103_104_NODDI_EddyCorrected.eddy_rotated_bvecs -o tractSeg --raw_diffusion_input --csd_type csd_msmt_5tt --brain_mask 103_104_NODDI_EddyCorrected_b0MeanMask.nii.gz --preprocess --output_type tract_segmentation

Any thoughts?

Thanks so much!