Abstract:Pancreatic ductal adenocarcinoma (PDAC) segmentation on contrast-enhanced CT is inherently ambiguous: inter-rater disagreement among experts reflects genuine uncertainty rather than annotation noise. Standard deep learning approaches assume a single ground truth, producing probabilistic outputs that can be poorly calibrated and difficult to interpret under such ambiguity. We present TwinTrack, a framework that addresses this gap through post-hoc calibration of ensemble segmentation probabilities to the empirical mean human response (MHR) -the fraction of expert annotators labeling a voxel as tumor. Calibrated probabilities are thus directly interpretable as the expected proportion of annotators assigning the tumor label, explicitly modeling inter-rater disagreement. The proposed post-hoc calibration procedure is simple and requires only a small multi-rater calibration set. It consistently improves calibration metrics over standard approaches when evaluated on the MICCAI 2025 CURVAS-PDACVI multi-rater benchmark.
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2604.15950 [cs.LG] |
| (or arXiv:2604.15950v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2604.15950 arXiv-issued DOI via DataCite (pending registration) |
From: sylvain faisan [view email] [via CCSD proxy]
[v1]
Fri, 17 Apr 2026 11:11:43 UTC (2,872 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。