Atlas is Your Perfect Context: One-Shot Customization for Generalizable Foundational Medical Image Segmentation
For clinicians and researchers needing accurate segmentation in underrepresented clinical contexts, AtlasSegFM offers a lightweight, deployable solution that adapts foundation models with a single annotated example.
AtlasSegFM enables one-shot customization of medical image segmentation foundation models to new clinical contexts using an atlas-guided framework, achieving consistent improvements across six datasets, especially on small structures.
Accurate segmentation of anatomical structures in medical images is essential for diagnosis and treatment planning. While recent interactive segmentation foundation models enhance generalization through large-scale multimodal pretraining, they still depend on precise prompts and can fail in underrepresented clinical contexts (e.g., small organs-at-risk). We present AtlasSegFM, an atlas-guided framework that customizes off-the-shelf foundation models to new clinical contexts with a single annotated example. AtlasSegFM 1) performs atlas-query registration to generate context-aware prompts, 2) refines the segmentation with a frozen foundation model, and 3) applies a lightweight adaptive fusion module to combine atlas priors with foundation-model inputs and predictions. Extensive experiments on six public and in-house datasets across radiotherapy and vascular scenarios show consistent gains, with the largest improvements on small and delicate structures. AtlasSegFM provides a lightweight, deployable solution for one-shot customization of segmentation foundation models in real-world clinical workflows.