# Tractography QC Meeting

Sept 16, 2021

## Discussion Items

- how to do QC?
- summer students viewed example anatomical tracts from the [ORG Atlas](https://www.youtube.com/watch?v=aINJeYvIPcE) and compared to subject-specific tracts
- standardizing QC notes
- OPT: 1 (good), 2 (minor issues), 3 (major issue)
- SPINS/SPIN-ASD: Missing, Poor (minor issue), Error (major issue) - biologically unfeasible tract
- exclusion criteria/cutoffs based on visual or quantitative QC - might be dataset specific
- when to exclude subject/tract?
- OPT: Denise's QC on major issues might have been too strict; excluding all "3" ratings would remove too much
- SPINS: if participant missing/poor n tracts (maybe 10), automatically excluded; if missing/poor more than 2 tracts of interest, also excluded
- when to try re-running subject?
- are there specific tracts that fail across studies and might not be reliable?
- should we correct tracts output by whitematteranalysis
- filtering when too many fibers present - SPINS: filtered fiber count being too high/low beyond 4 sd
- 
- introducing waypoints or ROIs to remove fibers from a specific region - pair WMA with small corrections to the data as required
- 
- other measures to consider
- bundle profiles along tract
- 
- ROI Atlases to use
- [easy lausanne](https://github.com/mattcieslak/easy_lausanne) breaks up DKT atlas into smaller chunks
- interpolating missing data
- interpret on the basis of other proximal tracts (Catani paper)

## Action Items
- check correspondence between eddy_quad excluded subjects and wma excluded subjects
- calculate fd from eddy_quad metrics
- get subjects that need to be re-run
-