Generalizable Single-Source Cross-modality Medical Image Segmentation via Invariant Causal Mechanisms
This addresses domain shifts in medical imaging, enabling models trained on a single source to generalize better across unseen modalities, which is incremental but practical for healthcare applications.
The paper tackled the problem of single-source domain generalization for cross-modality medical image segmentation by combining causality-inspired insights with diffusion-based augmentation, resulting in consistent outperformance of state-of-the-art methods across three anatomies and imaging modalities.
Single-source domain generalization (SDG) aims to learn a model from a single source domain that can generalize well on unseen target domains. This is an important task in computer vision, particularly relevant to medical imaging where domain shifts are common. In this work, we consider a challenging yet practical setting: SDG for cross-modality medical image segmentation. We combine causality-inspired theoretical insights on learning domain-invariant representations with recent advancements in diffusion-based augmentation to improve generalization across diverse imaging modalities. Guided by the ``intervention-augmentation equivariant'' principle, we use controlled diffusion models (DMs) to simulate diverse imaging styles while preserving the content, leveraging rich generative priors in large-scale pretrained DMs to comprehensively perturb the multidimensional style variable. Extensive experiments on challenging cross-modality segmentation tasks demonstrate that our approach consistently outperforms state-of-the-art SDG methods across three distinct anatomies and imaging modalities. The source code is available at \href{https://github.com/ratschlab/ICMSeg}{https://github.com/ratschlab/ICMSeg}.