Double-branch U-Net for multi-scale organ segmentation

doi:10.1016/j.ymeth.2022.07.002

	Double-branch U-Net for multi-scale organ segmentation
	Liu, Yuhao 1,2; Qin, Caijie 1,3; Yu, Zhiqian 4; Yang, Ruijie 5; Suqing, Tian 5; Liu, Xia 2; Ma, Xibo1,6
Source Publication	METHODS
ISSN	1046-2023
	2022-09-01
Volume	205 Pages:220-225
Corresponding Author	Liu, Xia(liuxia@hrbust.edu.cn) ; Ma, Xibo(xibo.ma@nlpr.ia.ac.cn)
Abstract	U-Net has achieved great success in the task of medical image segmentation. It encodes and extracts information from several convolution blocks, and then decodes the feature maps to get the segmentation results. Our ex-periments show that in a multi-scale medical segmentation task, excessive downsampling will cause the model to ignore the small segmentation objects and thus fail to complete the segmentation task. In this work, we propose a more complete method Double-branch U-Net (2BUNet) to solve the multi-scale organ segmentation challenge. Our model is divided into four parts: main branch, tributary branch, information exchange module and classi-fication module. The main advantages of the new model consist of: (1) Extracting information to improve model decoding capabilities using the complete encoding structure. (2) The information exchange module is added to the main branch and tributaries to provide regularization for the model, so as to avoid the large gap between the two paths. (3) Main branch structure for extracting major features of large organ. (4) The tributary structure is used to enlarge the image to extract the microscopic characteristics of small organ. (5) A classification assistant module is proposed to increase the class constraint for the output tensor. The comparative experiments show that our method achieves state-of-the-art performances in real scenes.
Keyword	Multi -scale segmentation Organ segmentation Medical image
DOI	10.1016/j.ymeth.2022.07.002
Indexed By	SCI
Language	英语
Funding Project	National Key Research programs of China[2016YFA0100900] ; National Key Research programs of China[2016YFA0100902] ; Chinese National Natural Science Foundation[82090051] ; Chinese National Natural Science Foundation[81871442] ; Youth Innovation Promotion Association CAS[Y201930]
Funding Organization	National Key Research programs of China ; Chinese National Natural Science Foundation ; Youth Innovation Promotion Association CAS
WOS Research Area	Biochemistry & Molecular Biology
WOS Subject	Biochemical Research Methods ; Biochemistry & Molecular Biology
WOS ID	WOS:000835183700002
Publisher	ACADEMIC PRESS INC ELSEVIER SCIENCE
Citation statistics
Document Type	期刊论文
Identifier	http://ir.ia.ac.cn/handle/173211/49801
Collection	模式识别国家重点实验室_生物识别与安全技术
Corresponding Author	Liu, Xia; Ma, Xibo
Affiliation	1.Chinese Acad Sci, Inst Automat, Beijing 100190, Peoples R China 2.Harbin Univ Sci & Technol, Harbin 150080, Peoples R China 3.Sanming Univ, Inst Informat Engn, Sanming 365004, Peoples R China 4.Northwest Univ, Informat Sci & Technol, Xian, Peoples R China 5.Peking Univ Third Hosp, Dept Radiat Oncol, Beijing 100190, Peoples R China 6.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China
First Author Affilication	Institute of Automation, Chinese Academy of Sciences
Corresponding Author Affilication	Institute of Automation, Chinese Academy of Sciences
Recommended Citation GB/T 7714	Liu, Yuhao,Qin, Caijie,Yu, Zhiqian,et al. Double-branch U-Net for multi-scale organ segmentation[J]. METHODS,2022,205:220-225.
APA	Liu, Yuhao.,Qin, Caijie.,Yu, Zhiqian.,Yang, Ruijie.,Suqing, Tian.,...&Ma, Xibo.(2022).Double-branch U-Net for multi-scale organ segmentation.METHODS,205,220-225.
MLA	Liu, Yuhao,et al."Double-branch U-Net for multi-scale organ segmentation".METHODS 205(2022):220-225.

Files in This Item:
There are no files associated with this item.

If you have any objections to this item, please fill out the form below and the administrator will contact you as soon as possible.
Content:
Email：	*
Institution:
Verification Code:	Refresh

Any comments and suggestions are welcomed.
Title:	*
Content:
Email：	*
Verification Code:	Refresh

模式识别国家重点实验室知识库