Dense Depth Completion Based on Multi-Scale Confidence and Self-Attention Mechanism for Intestinal Endoscopy

doi:10.1109/ICRA48891.2023.10161549

科研成果详情

题名	Dense Depth Completion Based on Multi-Scale Confidence and Self-Attention Mechanism for Intestinal Endoscopy
作者	Liu, Ruyu 1,2; Liu, Zhengzhe 1; Zhang, Haoyu 1; Zhang, Guodao 3; Zuo, Zhigui4; Sheng, Weiguo 1
发表日期	2023
会议录名称	Proceedings - IEEE International Conference on Robotics and Automation 影响因子和分区
语种	英语
原始文献类型	Conference article (CA) ; Conference Paper
关键词	Deep learning Human robot interaction Man machine systems Surgery Textures 3-D shape 3d locations Attention mechanisms Depth completion Depth Estimation Depth information Depth sensors Humans-robot interactions Multi-scales Self-attention mechanism depth completion Endoscopy human-robot interaction self-attention mechanism
摘要	Doctors perform limited one-way intestine endoscopy, in which advanced surgical robots with depth sensors, such as stereo and ToF endoscopes, can only provide sparse and incomplete depth information. However, dense, accurate and instant depth estimation during endoscopy is vital for doctors to judge the 3D location and shape of intestinal tissues, which affects the human-robot interaction between doctors and surgical robots, such as the operation on the subsequent moving of the probe. In this paper, we present a deep learning-based dense depth completion method for intestine endoscopy. We utilize the scattered depth information from depth sensors to make up for the deficiency of features in the intestine and design a multi-scale confidence prediction network to extract dense geometric depth features. Then, we introduce the structure awareness module based on the self-attention mechanism in the depth completion network to enhance the geometry and texture features of the intestine. We also present a virtual multi-modal RGBD intestine dataset and conduct comprehensive experiments on a total of three intestine datasets. The experimental results clearly demonstrate that our method achieves better results in all metrics in all intestinal environments compared to state-of-the-art methods. © 2023 IEEE.
资助项目	National Natural Science Foundation of China[62202137];Natural Science Foundation of Zhejiang Province[LQ22F030004]
出版者	Institute of Electrical and Electronics Engineers Inc.
出版地	NEW YORK
ISSN	1050-4729
EISSN	2577-087X
卷号	2023-May
页码	7476-7482
DOI	10.1109/ICRA48891.2023.10161549
页数	7
WOS类目	Automation & Control Systems ; Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic ; Robotics
WOS研究方向	Automation & Control Systems ; Computer Science ; Engineering ; Robotics
WOS记录号	WOS:001048371100067
收录类别	EI ; CPCI ; SCOPUS
EI入藏号	20233514632062
EI主题词	Endoscopy
EI分类号	461.4 Ergonomics and Human Factors Engineering ; 461.6 Medicine and Pharmacology ; 731.5 Robotics
URL	查看原文
Scopus记录号	2-s2.0-85168708440
通讯作者地址	[Sheng, Weiguo]Hangzhou Normal Univ, Sch Informat Sci & Technol, Hangzhou 311121, Zhejiang, Peoples R China.
Scopus学科分类	Software;Control and Systems Engineering;Electrical and Electronic Engineering;Artificial Intelligence
会议名称	2023 IEEE International Conference on Robotics and Automation, ICRA 2023
会议地点	London, United kingdom
会议日期	May 29, 2023 - June 2, 2023
引用统计
文献类型	会议论文
条目标识符	https://kms.wmu.edu.cn/handle/3ETUA0LF/183122
专题	附属第一医院_肛肠外科
通讯作者	Sheng, Weiguo
作者单位	1.Hangzhou Normal Univ, Sch Informat Sci & Technol, Hangzhou 311121, Zhejiang, Peoples R China; 2.Chinese Acad Sci, Haixi Inst, Quanzhou Inst Equipment Mfg, Quanzhou 362000, Peoples R China; 3.Hangzhou Dianzi Univ, Dept Digital Media Technol, Hangzhou 310018, Peoples R China; 4.Wenzhou Med Univ, Dept Colorectal Surg, Affiliated Hosp 1, Wenzhou 325035, Peoples R China
推荐引用方式 GB/T 7714	Liu, Ruyu,Liu, Zhengzhe,Zhang, Haoyu,et al. Dense Depth Completion Based on Multi-Scale Confidence and Self-Attention Mechanism for Intestinal Endoscopy[C]. NEW YORK:Institute of Electrical and Electronics Engineers Inc.,2023:7476-7482.