DFBU-Net: Double-branch flat bottom U-Net for efficient medical image segmentation

作者全名："Yin, Hao; Wang, Yi; Wen, Jing; Wang, Guangxian; Lin, Bo; Yang, Weibin; Ruan, Jian; Zhang, Yi"

作者地址："[Yin, Hao; Wang, Yi; Wen, Jing] Chongqing Univ, Sch Comp Sci, Chongqing 400000, Peoples R China; [Wang, Guangxian] Chongqing Med Univ, Banan Hosp, Dept Radiol, Chongqing 400000, Peoples R China; [Lin, Bo; Yang, Weibin; Ruan, Jian; Zhang, Yi] Chongqing Univ, Intelligent Oncol Res Ctr, Canc Hosp, Chongqing 400000, Peoples R China"

通信作者："Wang, Y (通讯作者)，Chongqing Univ, Sch Comp Sci, Chongqing 400000, Peoples R China."

来源：BIOMEDICAL SIGNAL PROCESSING AND CONTROL

ESI学科分类：ENGINEERING

WOS号：WOS:001137517100001

JCR分区：Q1

影响因子：4.9

年份：2024

卷号：90

期号：　

开始页：　

结束页：　

文献类型：Article

关键词：U-Net; Medical image segmentation; Medical characteristic; Double-branch

摘要："In the field of medical image processing, segmenting tissues and organs in CT/MRI and other medical sequence images is a vital yet challenging task. Analyzing the MICCAI competition, we have identified two problems in current methods for medical image organ segmentation: (1) There is a bottleneck in organ segmentation, with marginal room for improvement, as algorithmic capabilities have already surpassed the task's inherent difficulty. (2) Most current research focuses on stacking and enhancing new modules for segmentation while overlooking the inherent characteristics of medical sequence images. To overcome these two problems, firstly, we have encapsulated the three characteristics of CT/MRI medical sequence image segmentation: semantic correctness, edge accuracy, and 3D structure. Secondly, we delved into the most information-rich downsampling stage in terms of detail and semantics. Subsequently, we designed a flat-bottom double-branch network (DFBU-Net) based on the U-Net architecture. The high-resolution flat bottom branch of this network maintained a 1/4 feature map size to ensure the preservation of rich detail information, while the low-resolution branch underwent progressive downsampling to capture more semantic information. To prevent information loss, cross-fusion was performed at each stage of the model's two branches. Finally, DFBU-Net was evaluated on the MICCAI FLARE2021 dataset (DSC:93.61%, NSD:85.01%). Particularly, in the challenging task of pancreatic segmentation, our model outperformed the first-place model by 0.72% in DSC and 2.92% in NSD. Furthermore, in the MICCAI PARSE2022 competition, DFBU-Net ranked ninth with a DICE score of 79.28%, demonstrating its excellent segmentation performance and generalization ability."

基金机构："National Natural Science Foundation of China [61703062, 61906022]; Chongqing medical scientific re-search project (Joint project of Chongqing Health Commission and Science and Technology Bureau) [2020MSXM088]; Fundamental Research Funds for the Central Universities, China [2022CDJYGRH-015]; Medical Scientific Research Project of Chongqing Medical and Health Committee [2018GDRC006]; Humanities and Social Science Planning Fund from the Ministry of Education, China [21YJAZH013]"

基金资助正文："This work was supported by National Natural Science Foundation of China (No. 61703062, 61906022), Chongqing medical scientific research project (Joint project of Chongqing Health Commission and Science and Technology Bureau) (No. 2020MSXM088), the Fundamental Research Funds for the Central Universities, China (No. 2022CDJYGRH-015), Medical Scientific Research Project of Chongqing Medical and Health Committee (No. 2018GDRC006), and the Humanities and Social Science Planning Fund from the Ministry of Education, China (No. 21YJAZH013)."