Set-Inclusive Uncertainty Modeling for Robust Brain Tumor Segmentation

2026-06-29

AI Summary

This work addresses the challenge of unreliable and overconfident predictions in multimodal MRI brain tumor segmentation when certain imaging modalities are missing during inference, leading to significant information loss. To mitigate this issue, the authors propose a probabilistic representation framework that models features as Gaussian distributions, where the mean encodes task-relevant information and the variance explicitly quantifies uncertainty induced by missing modalities. By incorporating variance regularization and set-inclusion constraints, the method enforces hierarchical consistency in uncertainty estimates across different modality subsets and effectively aligns multimodal information. Experimental results on the BraTS 2018 and 2020 datasets demonstrate that the proposed approach significantly outperforms existing baselines under various modality-missing scenarios, achieving more reliable and adaptive segmentation performance.

Abstract

Multimodal MRI is essential for accurate brain tumor segmentation. However, acquiring all modalities at inference is often challenging in practice, which causes intrinsic uncertainty due to unavoidable information loss. Without modeling this uncertainty, existing methods encode incomplete evidence into deterministic representations that appear plausible but lack reliability. In this regime, we propose a probabilistic representation framework that models representations as Gaussian distributions, where their mean captures task information and their variance measures uncertainty from missing evidence. To make variance reflect information deficiency, we regularize the mean from each partial configuration toward its full-modality counterpart, while scaling the variance with the discrepancy between their aligned means. We further introduce a set-inclusive strategy that exploits the hierarchical structure of modality subsets and enforces an ordering constraint to maintain their consistent uncertainty relationships. Extensive experiments on BraTS 2018 and 2020 demonstrate that our approach offers superior performance over baselines across diverse missing-modality scenarios. Code and model checkpoint are available at https://github.com/atlas-sky/SIUM.

Problem

• brain tumor segmentation
• missing modalities
• uncertainty modeling
• multimodal MRI
• information deficiency

Innovation

• set-inclusive uncertainty
• probabilistic representation
• missing-modality robustness
• Gaussian embedding
• multimodal MRI segmentation