I recently ran the longitudinal TRACULA pipeline in a large dataset (>600 participants) using Freesurfer v7.3.2. Each participant in this dataset has two time points of data. I've finished running trac-all -path, so each dataset has 42 reconstructed WM tracts. 

I see in the tutorial that visual QC is recommended to check path outputs for partial reconstruction errors. Given our large dataset (1200 scans x 42 tracts), I was wondering what QC procedures you would recommend. I know that visual inspection is the gold standard, but are there any automatic or semi-automatic procedures you would recommend?

Thank you,

Jenna