UniAda: Domain Unifying and Adapting Network for Generalizable Medical Image Segmentation

作     者：Zhang, Zhongzhou Chen, Yingyu Yu, Hui Wang, Zhiwen Wang, Shanshan Fan, Fenglei Shan, Hongming Zhang, Yi 

作者机构：Sichuan Univ Coll Comp Sci Chengdu 610065 Peoples R China Sichuan Univ Minist Educ Key Lab Data Protect & Intelligent Management Chengdu 610207 Peoples R China Anhui Univ Minist Educ Inst Phys Sci & Informat Technol Key Lab Intelligent Comp & Signal Proc Hefei 230601 Peoples R China Chinese Univ Hong Kong Dept Math Hong Kong Peoples R China Fudan Univ MOE Frontiers Ctr Brain Sci Inst Sci & Technol Brain Inspired Intelligence Shanghai 200433 Peoples R China Fudan Univ Key Lab Computat Neurosci & Brain Inspired Intelli Shanghai 200433 Peoples R China Shanghai Ctr Brain Sci & Brain Inspired Technol Shanghai 201210 Peoples R China Sichuan Univ Sch Cyber Sci & Engn Chengdu 610207 Peoples R China 

出 版 物：《IEEE TRANSACTIONS ON MEDICAL IMAGING》 (IEEE Trans. Med. Imaging)

年 卷 期：2025年第44卷第5期

页      面：1988-2001页

核心收录：

学科分类：0831[工学-生物医学工程（可授工学、理学、医学学位）] 0808[工学-电气工程] 1002[医学-临床医学] 08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 1009[医学-特种医学] 

基　　金：National Natural Science Foundation of China [62271335, 62471148] Sichuan Science and Technology Program [2021JDJQ0024] Sichuan University "From 0 to 1" Innovative Research Program [2022SCUH0016] 

主　　题：Adaptation models Training Image segmentation Testing Data models Uncertainty Perturbation methods Medical diagnostic imaging Metalearning Representation learning Domain generalization medical image segmentation base model unifying and adapting 

摘      要：Learning a generalizable medical image segmentation model is an important but challenging task since the unseen (testing) domains may have significant discrepancies from seen (training) domains due to different vendors and scanning protocols. Existing segmentation methods, typically built upon domain generalization (DG), aim to learn multi-source domain-invariant features through data or feature augmentation techniques, but the resulting models either fail to characterize global domains during training or cannot sense unseen domain information during testing. To tackle these challenges, we propose a domain Unifying and Adapting network (UniAda) for generalizable medical image segmentation, a novel unifying while training, adapting while testing paradigm that can learn a domain-aware base model during training and dynamically adapt it to unseen target domains during testing. First, we propose to unify the multi-source domains into a global inter-source domain via a novel feature statistics update mechanism, which can sample new features for the unseen domains, facilitating the training of a domain base model. Second, we leverage the uncertainty map to guide the adaptation of the trained model for each testing sample, considering the specific target domain may be outside the global inter-source domain. Extensive experimental results on two public cross-domain medical datasets and one in-house cross-domain dataset demonstrate the strong generalization capacity of the proposed UniAda over state-of-the-art DG methods. The source code of our UniAda is available at https://***/ZhouZhang233/UniAda.