AC-MIL: Weakly Supervised Atrial LGE-MRI Quality Assessment via Adversarial Concept Disentanglement

2026-04-11

AI Summary

This work addresses the challenge of poor image quality in atrial late gadolinium enhancement MRI (LGE-MRI), which arises from motion artifacts, irregular respiration, and suboptimal acquisition timing, and for which existing methods lack interpretable failure-mode feedback. The authors propose the AC-MIL framework, which, under only volumetric-level weak supervision, decomposes global image quality into clinically defined radiological concepts. By integrating adversarial concept disentanglement, an unsupervised residual branch to prevent information leakage, and a spatial diversity constraint to decouple and localize attention maps, AC-MIL enables interpretable quality attribution within a weakly supervised multiple instance learning setting. This approach is the first to provide such interpretability while accurately localizing the causes of non-diagnostic scans on clinical datasets, all while maintaining ordinal grading performance on par with state-of-the-art methods.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Interpretability, Explainability, and Transparency
• Knowledge Representation and Reasoning: Qualitative Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

High-quality Late Gadolinium Enhancement (LGE) MRI can be helpful for atrial fibrillation management, yet scan quality is frequently compromised by patient motion, irregular breathing, and suboptimal image acquisition timing. While Multiple Instance Learning (MIL) has emerged as a powerful tool for automated quality assessment under weak supervision, current state-of-the-art methods map localized visual evidence to a single, opaque global feature vector. This black box approach fails to provide actionable feedback on specific failure modes, obscuring whether a scan degrades due to motion blur, inadequate contrast, or a lack of anatomical context. In this paper, we propose Adversarial Concept-MIL (AC-MIL), a weakly supervised framework that decomposes global image quality into clinically defined radiological concepts using only volume-level supervision. To capture latent quality variations without entangling predefined concepts, our framework incorporates an unsupervised residual branch guided by an adversarial erasure mechanism to strictly prevent information leakage. Furthermore, we introduce a spatial diversity constraint that penalizes overlap between distinct concept attention maps, ensuring localized and interpretable feature extraction. Extensive experiments on a clinical dataset of atrial LGE-MRI volumes demonstrate that AC-MIL successfully opens the MIL black box, providing highly localized spatial concept maps that allow clinicians to pinpoint the specific causes of non-diagnostic scans. Crucially, our framework achieves this deep clinical transparency while maintaining highly competitive ordinal grading performance against existing baselines. Code to be released on acceptance.

Problem

• Late Gadolinium Enhancement MRI
• Weakly Supervised Learning
• Image Quality Assessment
• Interpretability
• Multiple Instance Learning

Innovation

• Adversarial Concept Disentanglement
• Weakly Supervised Learning
• Multiple Instance Learning
• Interpretable AI
• LGE-MRI Quality Assessment