A patient-disjoint benchmark for chest CT segmentation demonstrates that slice-mixed evaluation inflates foreground Dice by 0.46 absolute (69% relative) compared to proper case-disjoint splits.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.IV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
CTSCAN: Evaluation Leakage in Chest CT Segmentation and a Reproducible Patient-Disjoint Benchmark
A patient-disjoint benchmark for chest CT segmentation demonstrates that slice-mixed evaluation inflates foreground Dice by 0.46 absolute (69% relative) compared to proper case-disjoint splits.