Salvato in:
Dettagli Bibliografici
Autori principali: Zhang, Ziyu, Yu, Yi, Zhu, Simeng, Aly, Ahmed, Gao, Yunhe, Gu, Ning, Xue, Yuan
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2512.18176
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909972267270144
author Zhang, Ziyu
Yu, Yi
Zhu, Simeng
Aly, Ahmed
Gao, Yunhe
Gu, Ning
Xue, Yuan
author_facet Zhang, Ziyu
Yu, Yi
Zhu, Simeng
Aly, Ahmed
Gao, Yunhe
Gu, Ning
Xue, Yuan
contents Accurate medical image segmentation is essential for clinical diagnosis and treatment planning. While recent interactive foundation models (e.g., nnInteractive) enhance generalization through large-scale multimodal pretraining, they still depend on precise prompts and often perform below expectations in contexts that are underrepresented in their training data. We present AtlasSegFM, an atlas-guided framework that customizes available foundation models to clinical contexts with a single annotated example. The core innovations are: 1) a pipeline that provides context-aware prompts for foundation models via registration between a context atlas and query images, and 2) a test-time adapter to fuse predictions from both atlas registration and the foundation model. Extensive experiments across public and in-house datasets spanning multiple modalities and organs demonstrate that AtlasSegFM consistently improves segmentation, particularly for small, delicate structures. AtlasSegFM provides a lightweight, deployable solution one-shot customization of foundation models in real-world clinical workflows. The code will be made publicly available.
format Preprint
id arxiv_https___arxiv_org_abs_2512_18176
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Atlas is Your Perfect Context: One-Shot Customization for Generalizable Foundational Medical Image Segmentation
Zhang, Ziyu
Yu, Yi
Zhu, Simeng
Aly, Ahmed
Gao, Yunhe
Gu, Ning
Xue, Yuan
Computer Vision and Pattern Recognition
Accurate medical image segmentation is essential for clinical diagnosis and treatment planning. While recent interactive foundation models (e.g., nnInteractive) enhance generalization through large-scale multimodal pretraining, they still depend on precise prompts and often perform below expectations in contexts that are underrepresented in their training data. We present AtlasSegFM, an atlas-guided framework that customizes available foundation models to clinical contexts with a single annotated example. The core innovations are: 1) a pipeline that provides context-aware prompts for foundation models via registration between a context atlas and query images, and 2) a test-time adapter to fuse predictions from both atlas registration and the foundation model. Extensive experiments across public and in-house datasets spanning multiple modalities and organs demonstrate that AtlasSegFM consistently improves segmentation, particularly for small, delicate structures. AtlasSegFM provides a lightweight, deployable solution one-shot customization of foundation models in real-world clinical workflows. The code will be made publicly available.
title Atlas is Your Perfect Context: One-Shot Customization for Generalizable Foundational Medical Image Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.18176