Semi-Supervised Medical Image Segmentation with Co-Distribution Alignment

被引:1
作者
Wang, Tao [1 ,2 ,3 ]
Huang, Zhongzheng [2 ]
Wu, Jiawei [4 ]
Cai, Yuanzheng [1 ]
Li, Zuoyong [1 ]
机构
[1] Minjiang Univ, Coll Comp & Control Engn, Fujian Prov Key Lab Informat Proc & Intelligent Co, Fuzhou 350108, Peoples R China
[2] Fuzhou Univ, Coll Comp & Data Sci, Fuzhou 350108, Peoples R China
[3] Wuyi Univ, Key Lab Cognit Comp & Intelligent Informat Proc, Fujian Educ Inst, Nanping 354300, Wuyishan, Peoples R China
[4] Fujian Agr & Forestry Univ, Sch Elect & Mech Engn, Fuzhou 350002, Peoples R China
来源
BIOENGINEERING-BASEL | 2023年 / 10卷 / 07期
基金
中国国家自然科学基金;
关键词
medical image segmentation; semi-supervised learning; distribution alignment; co-training;
D O I
10.3390/bioengineering10070869
中图分类号
Q81 [生物工程学(生物技术)]; Q93 [微生物学];
学科分类号
071005 ; 0836 ; 090102 ; 100705 ;
摘要
Medical image segmentation has made significant progress when a large amount of labeled data are available. However, annotating medical image segmentation datasets is expensive due to the requirement of professional skills. Additionally, classes are often unevenly distributed in medical images, which severely affects the classification performance on minority classes. To address these problems, this paper proposes Co-Distribution Alignment (Co-DA) for semi-supervised medical image segmentation. Specifically, Co-DA aligns marginal predictions on unlabeled data to marginal predictions on labeled data in a class-wise manner with two differently initialized models before using the pseudo-labels generated by one model to supervise the other. Besides, we design an over-expectation cross-entropy loss for filtering the unlabeled pixels to reduce noise in their pseudo-labels. Quantitative and qualitative experiments on three public datasets demonstrate that the proposed approach outperforms existing state-of-the-art semi-supervised medical image segmentation methods on both the 2D CaDIS dataset and the 3D LGE-MRI and ACDC datasets, achieving an mIoU of 0.8515 with only 24% labeled data on CaDIS, and a Dice score of 0.8824 and 0.8773 with only 20% data on LGE-MRI and ACDC, respectively.
引用
收藏
页数:20
相关论文
共 53 条
  • [31] Sajjadi M, 2016, ADV NEUR IN, V29
  • [32] Dynamic Joint Distribution Alignment Network for Bearing Fault Diagnosis Under Variable Working Conditions
    Shen, Changqing
    Wang, Xu
    Wang, Dong
    Li, Yongxiang
    Zhu, Jun
    Gong, Mingming
    [J]. IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, 2021, 70
  • [33] PROPERTIES OF CROSS-ENTROPY MINIMIZATION
    SHORE, JE
    JOHNSON, RW
    [J]. IEEE TRANSACTIONS ON INFORMATION THEORY, 1981, 27 (04) : 472 - 482
  • [34] Sohn K., 2020, Adv. Neural Inf. Process. Syst, P596, DOI DOI 10.5555/3495724.3495775
  • [35] Sun BC, 2016, AAAI CONF ARTIF INTE, P2058
  • [36] Sun Y, 2019, IEEE IMAGE PROC, P1535, DOI [10.1109/ICIP.2019.8803073, 10.1109/icip.2019.8803073]
  • [37] Tan MX, 2019, PR MACH LEARN RES, V97
  • [38] Tarvainen A, 2017, ADV NEUR IN, V30
  • [39] The journey to femtosecond laser-assisted cataract surgery: new beginnings or a false dawn?
    Trikha, S.
    Turnbull, A. M. J.
    Morris, R. J.
    Anderson, D. F.
    Hossain, P.
    [J]. EYE, 2013, 27 (04) : 461 - 473
  • [40] Focal Loss for Dense Object Detection
    Lin, Tsung-Yi
    Goyal, Priya
    Girshick, Ross
    He, Kaiming
    Dollar, Piotr
    [J]. 2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2017, : 2999 - 3007