[1]孙锐,麦华煜,李徵,等.松弛分布一致性的半监督医学图像分割[J].智能系统学报,2026,21(1):132-145.[doi:10.11992/tis.202507034]
SUN Rui,MAI Huayu,LI Zhi,et al.Relaxed distribution-wise consistency for semi-supervised medical image segmentation[J].CAAI Transactions on Intelligent Systems,2026,21(1):132-145.[doi:10.11992/tis.202507034]
点击复制
SUN Rui,MAI Huayu,LI Zhi,et al.Relaxed distribution-wise consistency for semi-supervised medical image segmentation[J].CAAI Transactions on Intelligent Systems,2026,21(1):132-145.[doi:10.11992/tis.202507034]
松弛分布一致性的半监督医学图像分割
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
21
期数:
2026年第1期
页码:
132-145
栏目:
学术论文—机器感知与模式识别
出版日期:
2026-03-05
- Title:
- Relaxed distribution-wise consistency for semi-supervised medical image segmentation
- Keywords:
- semi-supervised learning; medical images segmentation; consistency regularization; pseudo label; interpixel correlation; distribution-wise consistency; agent-based modeling; ranking alignment
- 分类号:
- TP391
- 摘要:
- 半监督医学图像分割可有效缓解医学图像标注成本高、效率低的问题，然而，医学图像中蕴含的丰富像素间相关性尚未被有效利用，影响了伪标签的可靠性。针对这一问题，本文分析了传统像素级一致性正则化方法在处理像素间结构关系时的局限性导致的性能瓶颈，提出一种融合像素间相关性的松弛分布一致性(relaxed distribution-wise consistency，RDC)方法。设计正交选择策略以选取代表性特征代理，通过分布一致性约束实现像素–代理相关性分布对齐，补充了传统像素级一致性监督；提出了排序对齐策略，松弛严格的分布数值对齐，从而提升了方法对噪声的鲁棒性。在ACDC、LA和Pancreas-NIH共3个公开数据集上的实验结果表明，RDC方法性能明显优于现有先进的半监督医学图像分割方法。本文研究结果可为半监督医学图像分割中未标注数据的利用策略设计提供新的思路与参考。
- Abstract:
- Semi-supervised medical image segmentation effectively alleviates the challenges of high annotation costs and inefficiencies in clinical settings. However, the rich interpixel correlations in medical images have not been fully exploited, which can reduce the accuracy and reliability of pseudo-labels used for training. To address this limitation, this work analyzes the limitations of conventional pixel-wise consistency regularization and proposes a relaxed distribution-wise consistency(RDC) method that integrates interpixel correlations. First, an orthogonal selection strategy is designed to construct representative feature agents. The distribution-wise consistency then aligns the pixel–agent correlations across augmentation views. Second, a ranking alignment strategy is employed to relax strict value-wise alignment, thereby increasing robustness to noise. Experiments on three challenging public datasets—ACDC, LA, and Pancreas-NIH—show that the proposed RDC method outperforms existing semi-supervised medical image segmentation methods. These findings provide new insights and references for designing effective strategies to exploit unlabeled data in semi-supervised medical image segmentation.
- 参考文献/References:
-
[1] HAIDEGGER T. Autonomy for surgical robots: concepts and paradigms[J]. IEEE transactions on medical robotics and bionics, 2019, 1(2): 65-76
[2] BIRLO M, ESWARDS P J E, CLARKSON M, et al. Utility of optical see-through head mounted displays in augmented reality-assisted surgery: a systematic review[J]. Medical image analysis, 2022, 77: 102361
[3] 方超伟, 李雪, 李钟毓, 等. 基于双模型交互学习的半监督医学图像分割[J]. 自动化学报, 2023, 49(4): 805-819 FANG ChaoWei, LI Xue, Li ZhongYu, et al. Interactive dual-model learning for semi-supervised medical image segmentation[J]. Acta automatica sinica, 2023, 49(4): 805-819
[4] 石军, 王天同, 朱子琦, 等. 基于深度学习的医学图像分割方法综述[J]. 中国图象图形学报, 2025, 30(6): 2161-2186 SHI Jun, WANG Tiantong, ZHU Ziqi, et al. Deep learning-based medical image segmentation methods[J]. Journal of image and graphics, 2025, 30(6): 2161-2186
[5] MAI Huayu, SUN Rui, WU Feng. Relaxed class-consensus consistency for semi-supervised semantic segmentation[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Philadelphia: AAAI, 2025.
[6] SUN Rui, MAI Huayu, LI Wangkai, et al. Towards unbiased learning in semi-supervised semantic segmentation[C]//The Thirteenth International Conference on Learning Representations. Singapore: PMLR, 2025.
[7] CHEN Xiaokang, YUAN Yuhui, ZENG Gang, et al. Semi-supervised semantic segmentation with cross pseudo supervision[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Virtual: IEEE, 2021.
[8] 姜婷, 袭肖明, 岳厚光. 基于分布先验的半监督FCM的肺结节分类[J]. 智能系统学报, 2017, 12(5): 729-734 JIANG Ting, XI Xiaoming, YUE Houguang. Classification of pulmonary nodules by semi-supervised FCM based on prior distribution[J]. CAAI transactions on intelligent systems, 2017, 12(5): 729-734
[9] CHI Hanyang, PANG Jian, ZHANG Bingfeng, et al. Adaptive bidirectional displacement for semi-supervised medical image segmentation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2024.
[10] WU Yicheng, XU Minfeng, GE Zongyuan, et al. Semi-supervised left atrium segmentation with mutual consistency training[C]//International Conference on Medical Image Computing and Computer Assisted Intervention. Strasbourg: MICCAI Society, 2021.
[11] 邱成健, 刘青山, 宋余庆, 等. 基于循环显著性校准网络的胰腺分割方法[J]. 自动化学报, 2022, 48(11): 2703-2717 QIU Chengjian, LIU Qingshan, SONG Yuqing, et al. Pancreas segmentation based on recurrent saliency calibration network[J]. Acta automatica sinica, 2022, 48(11): 2703-2717
[12] MAI Huayu, SUN Rui, ZHANG Tianzhu et al. Rankmatch: exploring the better consistency regularization for semi-supervised semantic segmentation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2024.
[13] KE Zhanghan, QIU Di, LI Kaican, et al. Guided collaborative training for pixel-wise semi-supervised learning[C]//European Conference on Computer Vision. Glasgow: ECVA, 2020.
[14] CAO Hu, WANG Yueyue, CHEN Joy, et al. Swin-UNet: UNet-like pure transformer for medical image segmentation[C]//European Conference on Computer Vision Workshops. Tel Aviv: ECVA, 2022.
[15] BAI Yunchao, CHEN Duowen, LI Qingli, et al. Bidirectional copy-paste for semi-supervised medical image segmentation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023.
[16] GR?NWALD P D. The minimum description length principle[M]. Massachusetts: MIT Press, 2007.
[17] GUO Chuan, PLEISS G, SUN Yu, et al. On calibration of modern neural networks[C]//International Conference on Machine Learning. Sydney: PMLR, 2017.
[18] SUN Rui, LI Yihao, ZHANG Tianzhu, et al. Lesion-aware transformers for diabetic retinopathy grading[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Virtual: IEEE, 2021.
[19] 赵凤, 张咪咪, 刘汉强. 区域信息驱动的多目标进化半监督模糊聚类图像分割算法[J]. 电子与信息学报, 2019, 41(5): 1106-1113 ZHAO Feng, ZHANG Mimi, LIU Hanqiang. Multi-objective evolutionary semi-supervised fuzzy clustering image segmentation motivated by region information[J]. Journal of electronics & information technology, 2019, 41(5): 1106-1113
[20] CHEN Xinjian, UDUPA J K, BAGCI U, et al. Medical image segmentation by combining graph cuts and oriented active appearance models[J]. IEEE transactions on image processing, 2012, 21(4): 2035-2046
[21] KHADIDOS A, SANCHEZ V, LI C T. Weighted level set evolution based on local edge features for medical image segmentation[J]. IEEE transactions on image processing, 2017, 26(4): 1979-1991
[22] 刘云, 陆承泽, 李仕杰, 等. 基于高效的多尺度特征提取的轻量级语义分割[J]. 计算机学报, 2022, 45(7): 1517-1528 LIU Yun, LU Chengze, LI Shijie, et al. Effcient multi-scale feature extraction for lightweight semantic segmentation[J]. Chinese journal of computers, 2022, 45(7): 1517-1528
[23] SUN Rui, MAI Huayu, ZHANG Tianzhu, et al. DAW: exploring the better weighting function for semi-supervised semantic segmentation[C]//Proceedings of the Advances in Neural Information Processing Systems. New Orleans: PMLR, 2023.
[24] MAI Huayu, SUN Rui, ZHANG Tianzhu, et al. Dualrel: semi-supervised mitochondria segmentation from a prototype perspective[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023.
[25] 夏平, 施宇, 雷帮军, 等. 复小波域混合概率图模型的超声医学图像分割[J]. 自动化学报, 2021, 47(1): 185-196 XIA Ping, SHI Yu, LEI Bangjun, et al. Ultrasound medical image segmentation based on hybrid probabilistic graphical model in complex-wavelet domain[J]. Acta automatica sinica, 2021, 47(1): 185-196
[26] RONNEBERGER O, FISCHER P, BROX T. U-Net: convolutional networks for biomedical image segmentation[C]//International Conference on Medical Image Computing and Computer Assisted Intervention. Munich: MICCAI Society, 2015.
[27] ISENSEE F, JAEGER P F, KOHL S A A, et al. nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation[J]. Nature methods, 2021, 18(2): 203-211
[28] XU Ju, LI Mengzhang, ZHU Zhanxing. Automatic data augmentation for 3D medical image segmentation[C]//International Conference on Medical Image Computing and Computer Assisted Intervention. Lima: MICCAI Society, 2020.
[29] ZHOU Zongwei, SIDDIQUEE M M R, TAJBAKHSH N, et al. UNet++: redesigning skip connections to exploit multiscale features in image segmentation[J]. IEEE transactions on medical imaging, 2019, 39(6): 1856-1867
[30] WANG Zhengyang, ZOU Na, SHEN Dinggang, et al. Non-local U-Nets for biomedical image segmentation[C]//Proceedings of the AAAI Conference on Artificial Intelligence. New York: AAAI, 2020.
[31] ?I?EK ?, ABDULKADIR A, LIENKMP S S, et al. 3D U-Net: learning dense volumetric segmentation from sparse annotation[C]//International Conference on Medical Image Computing and Computer Assisted Intervention. Athens: MICCAI Society, 2016.
[32] LIU Yuyan, TIAN Yu, CHEN Yuanhong, et al. Perturbed and strict mean teachers for semi-supervised semantic segmentation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022.
[33] KARIMI D, SALCUDEAN S E. Reducing the hausdorff distance in medical image segmentation with convolutional neural networks[J]. IEEE transactions on medical imaging, 2019, 39(2): 499-513
[34] WONG K C L, MORADI M, TANG H, et al. 3D segmentation with exponential logarithmic loss for highly unbalanced object sizes[C]//International Conference on Medical Image Computing and Computer Assisted Intervention. Granada: MICCAI Society, 2018.
[35] SUN Rui, MAI Huayu, LI Wangkai, et al. Beyond confidence: exploiting homogeneous pattern for semi-supervised semantic segmentation[C]//International Conference on Machine Learning. Vancouver: PMLR, 2025.
[36] YU Lequan, WANG Shujun, LI Xiaomeng, et al. Uncertainty-aware model for semi-supervised 3D left atrium segmentation[C]//International Conference on Medical Image Computing and Computer Assisted Intervention. Shenzhen: MICCAI Society, 2019.
[37] 苏逸欣, 肖志勇. 面向3维心脏MRI分割的半监督双任务交叉一致性约束网络[J]. 中国图象图形学报, 2023, 28(4): 1198-1211 SU Yixin, XIAO Zhiyong. Semi-supervised dual-task cross-consistency constraint network for 3D cardiac MRI segmentation[J]. Journal of image and graphics, 2023, 28(4): 1198-1211
[38] XIA Yingda, YANG Dong, YU Zhiding, et al. Uncertainty-aware multi-view co-training for semi-supervised medical image segmentation and domain adaptation[J]. Medical image analysis, 2020, 65: 101766
[39] BERNARD O, LALANDE A, ZOTTI C, et al. Deep learning techniques for automatic MRI cardiac multi-structures segmentation and diagnosis: is the problem solved?[J]. IEEE transactions on medical imaging, 2018, 37(11): 2514-2525
[40] XIONG Zhaohan, XIA Qing, HU Zhiqiang, et al. A global benchmark of algorithms for segmenting the left atrium from late gadolinium-enhanced cardiac magnetic resonance imaging[J]. Medical image analysis, 2021, 67: 101832
[41] ROTH H R, LU Le, FARAG A, et al. Deeporgan: multi-level deep convolutional networks for automated pancreas segmentation[C]//International Conference on Medical Image Computing and Computer Assisted Intervention. Munich: MICCAI Society, 2015.
[42] 李才子, 刘瑞强, 钟焕新, 等. 面向半监督颅内动脉瘤 CTA图像分割的权重感知自洽模型[J]. 计算机科学技术学报, 2023, 38(3): 674-685 LI Caizi, LIU Ruiqiang, ZHONG Huanxin, et al. Semi-supervised intracranial aneurysm segmentation from CTA images via weight-perceptual self-ensembling model[J]. Journal of computer science & technology, 2023, 38(3): 674-685
[43] LI Shuailin, ZHANG Chuyu, HE Xuming. Shape-aware semi-supervised 3D semantic segmentation for medical images[C]//International Conference on Medical Image Computing and Computer Assisted Intervention. Lima: MICCAI Society, 2020.
[44] LUO Xiang, CHEN Jieneng, SONG Tao, et al. Semi-supervised medical image segmentation through dual-task consistency[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Virtual: AAAI, 2021.
[45] LUO Xiangde, LIAO Wenjun, CHEN Jieneng, et al. Efficient semi-supervised gross target volume of nasopharyngeal carcinoma segmentation via uncertainty rectified pyramid consistency[C]//International Conference on Medical Image Computing and Computer Assisted Intervention. Strasbourg: MICCAI Society, 2021.
[46] 刘少鹏, 洪佳明, 梁杰鹏, 等. 面向医学图像分割的半监督条件生成对抗网络[J]. 软件学报, 2020, 31(8): 2588-2602 LIU Shaopeng, HONG Jiaming, LIANG Jiepeng, et al. Medical image segmentation using semi-supervised conditional generative adversarial nets[J]. Journal of software, 2020, 31(8): 2588-2602
[47] WU Yicheng, WU Zhonghua, WU Qianyi, et al. Exploring smoothness and class-separation for semi-supervised medical image segmentation[C]//International Conference on Medical Image Computing and Computer Assisted Intervention. Singapore: MICCAI Society, 2022.
[48] WANG Yongchao, XIAO Bin, BI Xiuli, et al. Mcf: Mutual correction framework for semi-supervised medical image segmentation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023.
[49] MIAO Juzheng, CHEN Cheng, LIU Feng, et al. Caussl: causality-inspired semi-supervised learning for medical image segmentation[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. Paris: IEEE, 2023.
[50] HUANG Huimin, HUANG Yawen, XIE Shiao, et al. Combinatorial CNN-transformer learning with manifold constraints for semi-supervised medical image segmentation[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Vancouver: AAAI, 2024.
[51] GUO Xiuzhen, YU Lianyuan, SHI Ji, et al. Diff-CL: a novel cross pseudo-supervision method for semi-supervised medical image segmentation[EB/OL]. (2025-03-12)[2025-04-20]. https://arxiv.org/abs/2503.09408.
[52] LIU Xinyu, LI Wuyang, YUAN Yixuan. Diffrect: latent diffusion label rectification for semi-supervised medical image segmentation[C]//International Conference on Medical Image Computing and Computer Assisted Intervention. Morocco: MICCAI Society, 2024.
- 相似文献/References:
-
[1]程显毅,巩向普.改进的模糊C-均值算法在医学图像分割中的应用[J].智能系统学报,2010,5(1):80.
CHENG Xian-yi,GONG Xiang-pu.An improved fuzzy Cmeans algorithm for segmentation of medical images[J].CAAI Transactions on Intelligent Systems,2010,5():80.
[2]李建元,周脚根,关佶红,等.谱图聚类算法研究进展[J].智能系统学报,2011,6(5):405.
LI Jianyuan,ZHOU Jiaogen,GUAN Jihong,et al.A survey of clustering algorithms based on spectra of graphs[J].CAAI Transactions on Intelligent Systems,2011,6():405.
[3]刘杨磊,梁吉业,高嘉伟,等.基于Tri-training的半监督多标记学习算法[J].智能系统学报,2013,8(5):439.[doi:10.3969/j.issn.1673-4785.201305033]
LIU Yanglei,LIANG Jiye,GAO Jiawei,et al.Semi-supervised multi-label learning algorithm based on Tri-training[J].CAAI Transactions on Intelligent Systems,2013,8():439.[doi:10.3969/j.issn.1673-4785.201305033]
[4]蒋新华,高晟,廖律超,等.半监督SVM分类算法的交通视频车辆检测方法[J].智能系统学报,2015,10(5):690.[doi:10.11992/tis.201406044]
JIANG Xinhua,GAO Sheng,LIAO Ljuchao,et al.Traffic video vehicle detection based on semi-supervised SVM classification algorithm[J].CAAI Transactions on Intelligent Systems,2015,10():690.[doi:10.11992/tis.201406044]
[5]张钢,谢晓珊,黄英,等.面向大数据流的半监督在线多核学习算法[J].智能系统学报,2014,9(3):355.[doi:10.3969/j.issn.1673-4785.201403067]
ZHANG Gang,XIE Xiaoshan,HUANG Ying,et al.An online multi-kernel learning algorithm for big data[J].CAAI Transactions on Intelligent Systems,2014,9():355.[doi:10.3969/j.issn.1673-4785.201403067]
[6]赵晓晓,周治平.结合稀疏表示与约束传递的半监督谱聚类算法[J].智能系统学报,2018,13(5):855.[doi:10.11992/tis.201703013]
ZHAO Xiaoxiao,ZHOU Zhiping.A semi-supervised spectral clustering algorithm combined with sparse representation and constraint propagation[J].CAAI Transactions on Intelligent Systems,2018,13():855.[doi:10.11992/tis.201703013]
[7]严菲,王晓栋.鲁棒的半监督多标签特征选择方法[J].智能系统学报,2019,14(4):812.[doi:10.11992/tis.201809017]
YAN Fei,WANG Xiaodong.A robust, semi-supervised, and multi-label feature selection method[J].CAAI Transactions on Intelligent Systems,2019,14():812.[doi:10.11992/tis.201809017]
[8]程龙,钱文彬,王映龙,等.弱标记不完备决策系统的增量式属性约简算法[J].智能系统学报,2020,15(6):1079.[doi:10.11992/tis.202001017]
CHENG Long,QIAN Wenbin,WANG Yinglong,et al.An incremental attribute reduction algorithm for incomplete decision system with weak labeling[J].CAAI Transactions on Intelligent Systems,2020,15():1079.[doi:10.11992/tis.202001017]
[9]刘威,王薪予,刘光伟,等.融合关系特征的半监督图像分类方法研究[J].智能系统学报,2022,17(5):886.[doi:10.11992/tis.202109022]
LIU Wei,WANG Xinyu,LIU Guangwei,et al.Semi-supervised image classification method fused with relational features[J].CAAI Transactions on Intelligent Systems,2022,17():886.[doi:10.11992/tis.202109022]
[10]吕佳,邱小龙.基于K-means聚类和特征空间增强的噪声标签深度学习算法[J].智能系统学报,2024,19(2):267.[doi:10.11992/tis.202303014]
LYU Jia,QIU Xiaolong.A noisy label deep learning algorithm based on K-means clustering and feature space augmentation[J].CAAI Transactions on Intelligent Systems,2024,19():267.[doi:10.11992/tis.202303014]
备注/Memo
收稿日期:2025-7-30。
基金项目:国家自然科学基金项目(62425117, 62401335).
作者简介:孙锐，助理研究员，主要研究方向为计算机视觉与机器学习，尤其聚焦于视频理解、半监督学习与多模态大模型。发表学术论文30余篇，其中CCF-A类期刊/会议20余篇。E-mail：issunrui@mail.ustc.edu.cn。;刘瑜，研究员，博士生导师，国家杰出青年基金获得者，中国青年科技奖获得者，主要研究方向为多模态数据智能融合。发表学术论文80余篇，获专利授权50余项，登记软件著作权20余项。E-mail：liuyu_thu@mail.tsinghua.edu.cn。;何友，中国工程院院士，兼任中国人工智能学会副理事长、中国航空学会名誉副理事长、中国指挥与控制学会监事长等。主要研究方向为信号检测、信息融合、智能技术与应用。以第一完成人获4项国家科技进步二等奖，荣获何梁何利基金科学与技术进步奖、“求是”工程奖、全国留学回国人员成就奖、山东省科学技术最高奖等。E-mail： heyou@mail.tsinghua.edu.cn。
通讯作者:刘瑜. E-mail：liuyu_thu@mail.tsinghua.edu.cn
更新日期/Last Update:
2026-01-05