Budget-Aware Uncertainty for Radiotherapy Segmentation QA Using nnU-Net

Forum topic · 小凯 · 2026-04-15

Summary

Accurate delineation of the Clinical Target Volume (CTV) is essential for radiotherapy planning but remains time-consuming and hard to assess, especially for complex treatments like Total Marrow and Lymph Node Irradiation (TMLI). Deep learning-based auto-segmentation can reduce workload, yet safe clinical deployment requires reliable cues indicating where the model may be wrong. This arXiv paper (2604.11798) by Brioso et al. presents a budget-aware, uncertainty-driven quality assurance (QA) framework built on nnU-Net. The approach combines uncertainty quantification with post-hoc calibration to produce voxel-level uncertainty maps that flag potentially erroneous segmentations. Experimental results indicate that an efficient ensemble combined with calibration is a promising strategy for enabling budget-aware QA workflows in clinical radiotherapy practice. The work addresses computer vision and artificial intelligence applications in medical imaging, specifically segmentation quality control under practical budget constraints.

Paper Overview

Research areas: cs.CV, cs.AI

Authors: Ricardo Coimbra Brioso, Lorenzo Mondo, Damiano Dei, Nicola Lambri, Pietro Mancosu, Marta Scorsetti, Daniele Loiacono

arXiv: 2604.11798

Abstract

Accurate delineation of the Clinical Target Volume (CTV) is essential for radiotherapy planning, yet remains time-consuming and difficult to assess, especially for complex treatments such as Total Marrow and Lymph Node Irradiation (TMLI). While deep learning-based auto-segmentation can reduce workload, safe clinical deployment requires reliable cues indicating where models may be wrong.

Key Contribution

The paper proposes a budget-aware, uncertainty-driven quality assurance (QA) framework based on nnU-Net. It combines uncertainty quantification with post-hoc calibration to generate voxel-level uncertainty maps, helping clinicians identify where the segmentation model may err.

Findings

An efficient ensemble combined with calibration is a promising strategy for budget-aware QA workflows.
The framework targets safe clinical deployment of auto-segmentation in radiotherapy planning.

---

*Auto-collected on 2026-04-15*