VoxSeP: semi-positive voxels assist self-supervised 3D medical segmentation

doi:10.1007/s00530-022-00977-9

	VoxSeP: semi-positive voxels assist self-supervised 3D medical segmentation
	Yang, Zijie 1,2; Xie, Lingxi 3; Zhou, Wei 3; Huo, Xinyue 4; Wei, Longhui 3; Lu, Jian 5; Tian, Qi 3; Tang, Sheng 1,2,6
	2022-07-23
发表期刊	MULTIMEDIA SYSTEMS
ISSN	0942-4962
页码	16
摘要	Medical image segmentation enjoys the advantage of understanding 3D contexts, but 3D networks are prone to over-fitting due to the limited amount of annotated data. This paper investigates self-supervised pre-training, i.e., making use of unlabeled medical data to initialize 3D segmentation networks. We build our system upon contrastive learning, where the dependence on positive and negative samples obstructs it from satisfying performance on medical image datasets with fewer samples. To alleviate this issue, we present a novel proxy task that takes advantage of the property of human body similarity in medical scans, and defines the sub-volumes from the same position of different cases as Semi-Positive samples. Pre-trained on a mixed dataset containing 1254 CT volumes, the proposed approach, VoxSeP, transfers well to 4 downstream datasets with 2 different backbone networks. On both fully supervised and semi-supervised fine-tuning, VoxSeP achieves favorable averaged improvements (2% and 4%), which surpass several state-of-the-art counterparts.
关键词	3D medical image segmentation Contrastive learning Self-supervised learning Unsupervised learning
DOI	10.1007/s00530-022-00977-9
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[61871004] ; Project of Chinese Academy of Sciences[E141020]
WOS研究方向	Computer Science
WOS类目	Computer Science, Information Systems ; Computer Science, Theory & Methods
WOS记录号	WOS:000829128600001
出版者	SPRINGER
引用统计	被引频次：2[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/19481
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Tang, Sheng
作者单位	1.Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Beijing 100049, Peoples R China 3.Huawei Cloud, Beijing, Peoples R China 4.Univ Sci & Technol China, Hefei, Peoples R China 5.Peking Univ Third Hosp, Dept Urol, Beijing, Peoples R China 6.Zhejiang Lab, Res Inst Intelligent Comp, Hangzhou, Peoples R China
推荐引用方式 GB/T 7714	Yang, Zijie,Xie, Lingxi,Zhou, Wei,et al. VoxSeP: semi-positive voxels assist self-supervised 3D medical segmentation[J]. MULTIMEDIA SYSTEMS,2022:16.
APA	Yang, Zijie.,Xie, Lingxi.,Zhou, Wei.,Huo, Xinyue.,Wei, Longhui.,...&Tang, Sheng.(2022).VoxSeP: semi-positive voxels assist self-supervised 3D medical segmentation.MULTIMEDIA SYSTEMS,16.
MLA	Yang, Zijie,et al."VoxSeP: semi-positive voxels assist self-supervised 3D medical segmentation".MULTIMEDIA SYSTEMS (2022):16.