OptiConn Docs¶
White matter tractography lacks a gold standard for parameter settings, and most publications offer little rationale for their choices—parameters are often selected arbitrarily or by convention. This becomes critical when deriving structural connectomes for graph-theoretic analyses, where parameter decisions directly influence network topology and derived measures. OptiConn addresses this gap by automating tractography parameter discovery through Bayesian optimization, validating selections via cross-validation bootstrap, and applying optimal parameters to produce analysis-ready connectivity outputs with a principled, data-driven foundation.
- For setup, see Installation.
- For day-to-day runs, see Workflows and Demos.
- For configuration details, see Configuration and Validation Notes.
- For background, see User Guide and Methods.
Quick links¶
- Bayesian + Apply demo:
python scripts/opticonn_demo.py --step all - Cross-validation demo (seeds from Bayes):
python scripts/opticonn_cv_demo.py --workspace demo_workspace_cv - DSI Studio download: https://github.com/frankyeh/DSI-Studio/releases
Affiliation
- MRI-Lab Graz
- Contact: karl.koschutnig@uni-graz.at