Mitigating Shortcut Learning via Feature Disentanglement in Medical Imaging: A Benchmark Study

2026-02-17

AI Summary

This study addresses the limited generalization of deep learning models in medical imaging, which often stems from reliance on spurious correlations—commonly referred to as shortcut learning—unrelated to the underlying diagnostic task. For the first time, the authors systematically evaluate the effectiveness of feature disentanglement approaches, such as adversarial learning and latent space dependency minimization, in mitigating shortcut learning on medical imaging datasets containing both natural and synthetically introduced confounders. By integrating data rebalancing with model disentanglement strategies, the proposed framework not only significantly improves classification performance and robustness under strong spurious correlations but also uncovers differences in representation quality that conventional classification metrics fail to capture. This work thus offers an efficient and scalable solution to enhance the generalization capability of medical imaging models.

Technology Category

• Machine Learning: Adversarial Learning & Robustness
• Computer Vision: Adversarial Attacks & Robustness
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Although deep learning models in medical imaging often achieve excellent classification performance, they can rely on shortcut learning, exploiting spurious correlations or confounding factors that are not causally related to the target task. This poses risks in clinical settings, where models must generalize across institutions, populations, and acquisition conditions. Feature disentanglement is a promising approach to mitigate shortcut learning by separating task-relevant information from confounder-related features in latent representations. In this study, we systematically evaluated feature disentanglement methods for mitigating shortcuts in medical imaging, including adversarial learning and latent space splitting based on dependence minimization. We assessed classification performance and disentanglement quality using latent space analyses across one artificial and two medical datasets with natural and synthetic confounders. We also examined robustness under varying levels of confounding and compared computational efficiency across methods. We found that shortcut mitigation methods improved classification performance under strong spurious correlations during training. Latent space analyses revealed differences in representation quality not captured by classification metrics, highlighting the strengths and limitations of each method. Model reliance on shortcuts depended on the degree of confounding in the training data. The best-performing models combine data-centric rebalancing with model-centric disentanglement, achieving stronger and more robust shortcut mitigation than rebalancing alone while maintaining similar computational efficiency.

Problem

• shortcut learning
• medical imaging
• confounding factors
• spurious correlations
• generalization

Innovation

• feature disentanglement
• shortcut learning
• medical imaging
• adversarial learning
• confounder mitigation