Segmentation Performance
Cross-source consensus on Segmentation Performance from 1 sources and 4 claims.
1 sources · 4 claims
Benefits
Comparisons
Evidence quality
Highlighted claims
- Across independent datasets, AG-TAL achieved all-artery Dice scores between 74.46% and 81.17%. — AG-TAL: Anatomically-Guided Topology-Aware Loss for Multiclass Segmentation of the Circle of Willis Using Large-Scale Multi-Center Datasets
- AG-TAL achieved the best average Dice across artery-size groups in cross-validation testing. — AG-TAL: Anatomically-Guided Topology-Aware Loss for Multiclass Segmentation of the Circle of Willis Using Large-Scale Multi-Center Datasets
- AG-TAL improved all-artery Dice by 1.05 percentage points over the nnResUNet baseline. — AG-TAL: Anatomically-Guided Topology-Aware Loss for Multiclass Segmentation of the Circle of Willis Using Large-Scale Multi-Center Datasets
- AG-TAL ranked first overall on clDice, HD95, beta0 error, and beta error. — AG-TAL: Anatomically-Guided Topology-Aware Loss for Multiclass Segmentation of the Circle of Willis Using Large-Scale Multi-Center Datasets