Dynamic graph consistency and self-contrast learning for semi-supervised medical image segmentation

被引:0
作者
Li, Gang [1 ]
Xie, Jinjie [1 ]
Zhang, Ling [1 ]
Cheng, Guijuan [1 ]
Zhang, Kairu [1 ]
Bai, Mingqi [1 ]
机构
[1] Taiyuan Univ Technol, Coll Software, Taiyuan, Peoples R China
关键词
Semi-supervised learning; Dynamic graph consistency; Self-contrast learning; Medical image segmentation;
D O I
10.1016/j.neunet.2024.107063
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Semi-supervised medical image segmentation endeavors to exploit a limited set of labeled data in conjunction with a substantial corpus of unlabeled data, with the aim of training models that can match or even exceed the efficacy of fully supervised segmentation models. Despite the potential of this approach, most existing semisupervised medical image segmentation techniques that employ consistency regularization predominantly focus on spatial consistency at the image level, often neglecting the crucial role of feature-level channel information. To address this limitation, we propose an innovative method that integrates graph convolutional networks with a consistency regularization framework to develop a dynamic graph consistency approach. This method imposes channel-level constraints across different decoders by leveraging high-level features within the network. Furthermore, we introduce a novel self-contrast learning strategy, which performs image-level comparison within the same batch and engages in pixel-level contrast learning based on pixel positions. This approach effectively overcomes traditional contrast learning challenges related to identifying positive and negative samples, reduces computational resource consumption, and significantly improves model performance. Our experimental evaluation on three distinct medical image segmentation datasets indicates that the proposed method demonstrates superior performance across a variety of test scenarios.
引用
收藏
页数:14
相关论文
共 50 条
  • [21] Dual consistency regularization with subjective logic for semi-supervised medical image segmentation
    Lu, Shanfu
    Yan, Ziye
    Chen, Wei
    Cheng, Tingting
    Zhang, Zijian
    Yang, Guang
    [J]. COMPUTERS IN BIOLOGY AND MEDICINE, 2024, 170
  • [22] Curriculum Consistency Learning and Multi-Scale Contrastive Constraint in Semi-Supervised Medical Image Segmentation
    Ding, Weizhen
    Li, Zhen
    [J]. BIOENGINEERING-BASEL, 2024, 11 (01):
  • [23] Contour-aware consistency for semi-supervised medical image segmentation
    Li, Lei
    Lian, Sheng
    Luo, Zhiming
    Wang, Beizhan
    Li, Shaozi
    [J]. BIOMEDICAL SIGNAL PROCESSING AND CONTROL, 2024, 89
  • [24] Semi-Supervised Learning With Fact-Forcing for Medical Image Segmentation
    Bui, Phuoc-Nguyen
    Le, Duc-Tai
    Bum, Junghyun
    Kim, Seongho
    Song, Su Jeong
    Choo, Hyunseung
    [J]. IEEE ACCESS, 2023, 11 : 99413 - 99425
  • [25] Co-Manifold learning for semi-supervised medical image segmentation
    Peiris, Himashi
    Chen, Zhaolin
    Egan, Gary
    Harandi, Mehrtash
    [J]. NEUROCOMPUTING, 2025, 639
  • [26] Semi-supervised medical image segmentation network based on mutual learning
    Sun, Junmei
    Wang, Tianyang
    Wang, Meixi
    Li, Xiumei
    Xu, Yingying
    [J]. MEDICAL PHYSICS, 2025, 52 (03) : 1589 - 1600
  • [27] Multi-consistency for semi-supervised medical image segmentation via diffusion models
    Chen, Yunzhu
    Liu, Yang
    Lu, Manti
    Fu, Liyao
    Yang, Feng
    [J]. PATTERN RECOGNITION, 2025, 161
  • [28] Semi-supervised Medical Image Segmentation with Strong/Weak Task-Aware Consistency
    Wang, Hua
    Liu, Linwei
    Lin, Yiming
    Hu, Jingfei
    Zhang, Jicong
    [J]. PATTERN RECOGNITION AND COMPUTER VISION, PRCV 2024, PT XIV, 2025, 15044 : 17 - 31
  • [29] Consistency-Guided Differential Decoding for Enhancing Semi-Supervised Medical Image Segmentation
    Zeng, Qingjie
    Xie, Yutong
    Lu, Zilin
    Lu, Mengkang
    Zhang, Jingfeng
    Xia, Yong
    [J]. IEEE TRANSACTIONS ON MEDICAL IMAGING, 2025, 44 (01) : 44 - 56
  • [30] Dual-scale enhanced and cross-generative consistency learning for semi-supervised medical image segmentation
    Gu, Yunqi
    Zhou, Tao
    Zhang, Yizhe
    Zhou, Yi
    He, Kelei
    Gong, Chen
    Fu, Huazhu
    [J]. PATTERN RECOGNITION, 2025, 158