Defacing / skull-strip verification gate
- Dominant language
- Python
- Stars
- 1
- Forks
- 4
- Avg merge
- 15h 39m
- Merged PRs (30d)
- 24
Description
# Defacing / skull-strip verification gate
## Goal
Gate processing on defacing verification so we never process — and publish —
undefaced data. Three parts:
1. **Tracking field.** Add a defacing-status field to the tracking TSV (e.g.
`priority-openneuro-datasets.csv`) so each dataset's defacing state is
recorded.
2. **Block on it.** Don't submit/process a dataset until defacing is verified;
if undefaced data is found, follow OpenNeuro's data-removal policy.
3. **Procedure.** Establish how we verify — Joe's metrics + defacing sheets;
jbwexler's `bids-mosaic` ().
## Why M2
Publishing undefaced data is a privacy violation you'd have to redo/remove →
squarely the redo-if-passes litmus.
## Automation
Automating the verification is a separate M4 stub — may not be possible.
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.