模式识别与人工智能
Friday, Apr. 4, 2025 Home      About Journal      Editorial Board      Instructions      Ethics Statement      Contact Us                   中文
Pattern Recognition and Artificial Intelligence  2024, Vol. 37 Issue (9): 755-769    DOI: 10.16451/j.cnki.issn1003-6059.202409001
Papers and Reports Current Issue| Next Issue| Archive| Adv Search |
Enhanced Target Domain Representation BasedUnsupervised Cross-Domain Medical Image Segmentation
LIU Kai1, LU Runuo1, ZHENG Xiaorou1, DONG Shoubin1
1. Guangdong Provincial Key Laboratory of Multimodal Big Data Intelligent Analysis, School of Computer Science and Engineering, South China University of Technology, Guangzhou 510641

Download: PDF (2275 KB)   HTML (1 KB) 
Export: BibTeX | EndNote (RIS)      
Abstract  Medical images produced by different imaging modality devices exhibit varying degrees of distribution differences. Unsupervised domain adaptation methods typically aim to generalize models trained in the source domain to the unlabeled target domain by minimizing these distribution differences and using shared features between the source and target domains for result prediction. However, they often neglect the private features of the target domain. To address this issue, a method for enhanced target domain representation based unsupervised cross-domain medical image segmentation(TreUCMIS) is proposed in this paper. First, TreUCMIS acquires common features through shared feature learning, and a target domain feature encoder is trained through image reconstruction to capture the complete features of the target domain. Second, unsupervised self-training of the target domain strengthens the shared characteristics of deep and shallow features. Finally, the predicted results obtained from the shared and complete features are aligned, enabling the model to utilize the complete features of the target domain for segmentation and thus improving the generalization in the target domain. Experiments on two medical image segmentation datasets involving bidirectional domain adaptation tasks with CT and MRI(abdominal and cardiac datasets) demonstrate the effectiveness and superiority of TreUCMIS.
Key wordsUnsupervised Domain Adaptation(UDA)      Medical Images      Semantic Segmentation      Adversarial Learning      Consistency Regularization     
Received: 15 August 2024     
ZTFLH: TP 391  
Fund:National Natural Science Foundation of China(No.61976239), Innovation Fund of High-End Scientific Research Institutions of Zhongshan City(No.2019AG031)
Corresponding Authors: DONG Shoubin, Ph.D., professor. Her research interests include high performance computing, bioinformatics and biocomputing, and natural language processing.   
About author:: ZHENG Xiaorou, Ph.D., assistant professor. Her research interests include hyperspectral remote sensing imagery, semi-supervised learning and cross-domain transfer learning. LU Runuo, Master student. Her research interests include computer vision and remote sensing imagery. LIU Kai, Master student. His research interests include deep learning, computer vision and image recognition.
Service
E-mail this article
Add to my bookshelf
Add to citation manager
E-mail Alert
RSS
Articles by authors
LIU Kai
LU Runuo
ZHENG Xiaorou
DONG Shoubin
Cite this article:   
LIU Kai,LU Runuo,ZHENG Xiaorou等. Enhanced Target Domain Representation BasedUnsupervised Cross-Domain Medical Image Segmentation[J]. Pattern Recognition and Artificial Intelligence, 2024, 37(9): 755-769.
URL:  
http://manu46.magtech.com.cn/Jweb_prai/EN/10.16451/j.cnki.issn1003-6059.202409001      OR     http://manu46.magtech.com.cn/Jweb_prai/EN/Y2024/V37/I9/755
Copyright © 2010 Editorial Office of Pattern Recognition and Artificial Intelligence
Address: No.350 Shushanhu Road, Hefei, Anhui Province, P.R. China Tel: 0551-65591176 Fax:0551-65591176 Email: bjb@iim.ac.cn
Supported by Beijing Magtech  Email:support@magtech.com.cn