Semi-Supervised Neuron Segmentation via Reinforced Consistency Learning

被引:43
作者
Huang, Wei [1 ,2 ]
Chen, Chang [1 ,2 ]
Xiong, Zhiwei [1 ,2 ]
Zhang, Yueyi [1 ,2 ]
Chen, Xuejin [1 ,2 ]
Sun, Xiaoyan [1 ,2 ]
Wu, Feng [1 ,2 ]
机构
[1] Univ Sci & Technol China, Dept Elect Engn & Informat Sci, Hefei 230027, Peoples R China
[2] Hefei Comprehens Natl Sci Ctr, Inst Artificial Intelligence, Hefei 230088, Peoples R China
基金
中国国家自然科学基金;
关键词
Neurons; Task analysis; Image segmentation; Data models; Perturbation methods; Training; Information filters; Neuron segmentation; deep learning; semi-supervised learning; electron microscopy images; ELECTRON-MICROSCOPY; RECONSTRUCTION;
D O I
10.1109/TMI.2022.3176050
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Emerging deep learning-based methods have enabled great progress in automatic neuron segmentation from Electron Microscopy (EM) volumes. However, the success of existing methods is heavily reliant upon a large number of annotations that are often expensive and time-consuming to collect due to dense distributions and complex structures of neurons. If the required quantity of manual annotations for learning cannot be reached, these methods turn out to be fragile. To address this issue, in this article, we propose a two-stage, semi-supervised learning method for neuron segmentation to fully extract useful information from unlabeled data. First, we devise a proxy task to enable network pre-training by reconstructing original volumes from their perturbed counterparts. This pre-training strategy implicitly extracts meaningful information on neuron structures from unlabeled data to facilitate the next stage of learning. Second, we regularize the supervised learning process with the pixel-level prediction consistencies between unlabeled samples and their perturbed counterparts. This improves the generalizability of the learned model to adapt diverse data distributions in EM volumes, especially when the number of labels is limited. Extensive experiments on representative EM datasets demonstrate the superior performance of our reinforced consistency learning compared to supervised learning, i.e., up to 400% gain on the VOI metric with only a few available labels. This is on par with a model trained on ten times the amount of labeled data in a supervised manner. Code is available at https://github.com/weih527/SSNS-Net.
引用
收藏
页码:3016 / 3028
页数:13
相关论文
共 50 条
[41]   Semi-supervised Medical Image Segmentation with Semantic Distance Distribution Consistency Learning [J].
Liu, Linhu ;
Tian, Jiang ;
Shi, Zhongchao ;
Fan, Jianping .
PATTERN RECOGNITION AND COMPUTER VISION, PRCV 2022, PT II, 2022, 13535 :323-335
[42]   Apple Leaf Spot Segmentation Model Based on Consistency Semi-supervised Learning [J].
Ding, Yongjun ;
Yang, Wentao ;
Zhao, Yilong .
Nongye Jixie Xuebao/Transactions of the Chinese Society for Agricultural Machinery, 2024, 55 (12) :314-321
[43]   Progressive Learning With Cross-Window Consistency for Semi-Supervised Semantic Segmentation [J].
Dang, Bo ;
Li, Yansheng ;
Zhang, Yongjun ;
Ma, Jiayi .
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2024, 33 :5219-5231
[44]   Semi-Supervised Video Object Segmentation Based on Local and Global Consistency Learning [J].
Liang, Huagang ;
Liu, Lihua ;
Bo, Ying ;
Zuo, Chao .
IEEE ACCESS, 2021, 9 :127293-127304
[45]   Pseudo Labeling based Consistency Learning for Semi-supervised Medical Image Segmentation [J].
Cheng, Yintao ;
Zhang, Manli ;
Yang, Rui ;
Kang, Guixia .
2024 11TH INTERNATIONAL CONFERENCE ON BIOMEDICAL AND BIOINFORMATICS ENGINEERING, ICBBE 2024, 2024, :46-52
[46]   Liver Segmentation with Semi-Supervised Learning [J].
Gao, Yonghui ;
Li, Xiaoxiao ;
Liu, Jingjing .
PROCEEDINGS OF THE 2015 4TH NATIONAL CONFERENCE ON ELECTRICAL, ELECTRONICS AND COMPUTER ENGINEERING ( NCEECE 2015), 2016, 47 :312-319
[47]   SEMI-SUPERVISED SEMANTIC SEGMENTATION CONSTRAINED BY CONSISTENCY REGULARIZATION [J].
Li, Xiaoqiang ;
He, Qin ;
Dai, Songmin ;
Wu, Pin ;
Tong, Weiqin .
2020 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO (ICME), 2020,
[48]   Semi-Supervised and Unsupervised Deep Visual Learning: A Survey [J].
Chen, Yanbei ;
Mancini, Massimiliano ;
Zhu, Xiatian ;
Akata, Zeynep .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2024, 46 (03) :1327-1347
[49]   Decoupled Consistency for Semi-supervised Medical Image Segmentation [J].
Chen, Faquan ;
Fei, Jingjing ;
Chen, Yaqi ;
Huang, Chenxi .
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION, MICCAI 2023, PT I, 2023, 14220 :551-561
[50]   FMixAugment for Semi-supervised Learning with Consistency Regularization [J].
Lin, Huibin ;
Wang, Shiping ;
Liu, Zhanghui ;
Xiao, Shunxin ;
Du, Shide ;
Guo, Wenzhong .
PATTERN RECOGNITION AND COMPUTER VISION, PRCV 2021, PT II, 2021, 13020 :127-139