Causally Guided Gaussian Perturbations for Out-Of-Distribution Generalization in Medical Imaging
This addresses the challenge of deploying reliable deep learning models in medical imaging where subtle distribution shifts are common, offering an interpretable approach for improved generalization.
The paper tackles out-of-distribution generalization in medical imaging by proposing Causally-Guided Gaussian Perturbations, a lightweight framework that injects spatially varying noise guided by causal masks to encourage reliance on relevant features, resulting in consistent performance gains on the Camelyon17 benchmark.
Out-of-distribution (OOD) generalization remains a central challenge in deploying deep learning models to real-world scenarios, particularly in domains such as biomedical images, where distribution shifts are both subtle and pervasive. While existing methods often pursue domain invariance through complex generative models or adversarial training, these approaches may overlook the underlying causal mechanisms of generalization.In this work, we propose Causally-Guided Gaussian Perturbations (CGP)-a lightweight framework that enhances OOD generalization by injecting spatially varying noise into input images, guided by soft causal masks derived from Vision Transformers. By applying stronger perturbations to background regions and weaker ones to foreground areas, CGP encourages the model to rely on causally relevant features rather than spurious correlations.Experimental results on the challenging WILDS benchmark Camelyon17 demonstrate consistent performance gains over state-of-the-art OOD baselines, highlighting the potential of causal perturbation as a tool for reliable and interpretable generalization.