Why Invariance is Not Enough for Biomedical Domain Generalization and How to Fix It
For biomedical image segmentation, DropGen provides a simple, architecture-agnostic method to improve robustness against domain shifts, addressing a key barrier to clinical deployment.
DropGen improves domain generalization in 3D biomedical image segmentation by combining source-domain intensities with foundation model representations, achieving strong gains in both fully supervised and few-shot settings across diverse shifts.
We present DropGen, a simple and theoretically-grounded approach for domain generalization in 3D biomedical image segmentation. Modern segmentation models degrade sharply under shifts in modality, disease severity, clinical sites, and other factors, creating brittle models that limit reliable deployment. Existing domain generalization methods rely on extreme augmentations, mixing domain statistics, or architectural redesigns, yet incur significant implementation overhead and yield inconsistent performance across biomedical settings. DropGen instead proposes a principled learning strategy with minimal overhead that leverages both source-domain image intensities and domain-stable foundation model representations to train robust segmentation models. As a result, DropGen achieves strong gains in both fully supervised and few-shot segmentation across a broad range of shifts in biomedical studies. Unlike prior approaches, DropGen is architecture- and loss-agnostic, compatible with standard augmentation pipelines, computationally lightweight, and tackles arbitrary anatomical regions. Our implementation is freely available at https://github.com/sebodiaz/DropGen.