DeepVinci: Organ and Tool Segmentation with Edge Supervision and a Densely Multi-Scale Pyramid Module for Robot-Assisted Surgery

<b>Background</b>: Automated surgical navigation can be separated into three stages: (1) organ identification and localization, (2) identification of the organs requiring further surgery, and (3) automated planning of the operation path and steps. With its ideal visual and operating syst...

Full description

Saved in:
Bibliographic Details
Main Authors: Li-An Tseng, Yuan-Chih Tsai, Meng-Yi Bai, Mei-Fang Li, Yi-Liang Lee, Kai-Jo Chiang, Yu-Chi Wang, Jing-Ming Guo
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Diagnostics
Subjects:
Online Access:https://www.mdpi.com/2075-4418/15/15/1917
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849239764113293312
author Li-An Tseng
Yuan-Chih Tsai
Meng-Yi Bai
Mei-Fang Li
Yi-Liang Lee
Kai-Jo Chiang
Yu-Chi Wang
Jing-Ming Guo
author_facet Li-An Tseng
Yuan-Chih Tsai
Meng-Yi Bai
Mei-Fang Li
Yi-Liang Lee
Kai-Jo Chiang
Yu-Chi Wang
Jing-Ming Guo
author_sort Li-An Tseng
collection DOAJ
description <b>Background</b>: Automated surgical navigation can be separated into three stages: (1) organ identification and localization, (2) identification of the organs requiring further surgery, and (3) automated planning of the operation path and steps. With its ideal visual and operating system, the da Vinci surgical system provides a promising platform for automated surgical navigation. This study focuses on the first step in automated surgical navigation by identifying organs in gynecological surgery. <b>Methods</b>: Due to the difficulty of collecting da Vinci gynecological endoscopy data, we propose DeepVinci, a novel end-to-end high-performance encoder–decoder network based on convolutional neural networks (CNNs) for pixel-level organ semantic segmentation. Specifically, to overcome the drawback of a limited field of view, we incorporate a densely multi-scale pyramid module and feature fusion module, which can also enhance the global context information. In addition, the system integrates an edge supervision network to refine the segmented results on the decoding side. <b>Results</b>: Experimental results show that DeepVinci can achieve state-of-the-art accuracy, obtaining dice similarity coefficient and mean pixel accuracy values of 0.684 and 0.700, respectively. <b>Conclusions</b>: The proposed DeepVinci network presents a practical and competitive semantic segmentation solution for da Vinci gynecological surgery.
format Article
id doaj-art-d602f57ae3e541f3b902a0830d0b0a5e
institution Kabale University
issn 2075-4418
language English
publishDate 2025-07-01
publisher MDPI AG
record_format Article
series Diagnostics
spelling doaj-art-d602f57ae3e541f3b902a0830d0b0a5e2025-08-20T04:00:50ZengMDPI AGDiagnostics2075-44182025-07-011515191710.3390/diagnostics15151917DeepVinci: Organ and Tool Segmentation with Edge Supervision and a Densely Multi-Scale Pyramid Module for Robot-Assisted SurgeryLi-An Tseng0Yuan-Chih Tsai1Meng-Yi Bai2Mei-Fang Li3Yi-Liang Lee4Kai-Jo Chiang5Yu-Chi Wang6Jing-Ming Guo7Department of Electrical Engineering, National Taiwan University of Science and Technology, Taipei 106335, TaiwanGraduate Institute of Applied Science and Technology, National Taiwan University of Science and Technology, Taipei 106335, TaiwanGraduate Institute of Biomedical Engineering, National Taiwan University of Science and Technology, Taipei 10607, TaiwanGraduate Institute of Biomedical Engineering, National Taiwan University of Science and Technology, Taipei 10607, TaiwanDepartment of Obstetrics and Gynecology, Tri-Service General Hospital, National Defense Medical Center, Taipei 114202, TaiwanDepartment of Nursing, Tri-Service General Hospital, National Defense Medical Center, Taipei 114, TaiwanDepartment of Obstetrics and Gynecology, Tri-Service General Hospital, National Defense Medical Center, Taipei 114202, TaiwanDepartment of Electrical Engineering, National Taiwan University of Science and Technology, Taipei 106335, Taiwan<b>Background</b>: Automated surgical navigation can be separated into three stages: (1) organ identification and localization, (2) identification of the organs requiring further surgery, and (3) automated planning of the operation path and steps. With its ideal visual and operating system, the da Vinci surgical system provides a promising platform for automated surgical navigation. This study focuses on the first step in automated surgical navigation by identifying organs in gynecological surgery. <b>Methods</b>: Due to the difficulty of collecting da Vinci gynecological endoscopy data, we propose DeepVinci, a novel end-to-end high-performance encoder–decoder network based on convolutional neural networks (CNNs) for pixel-level organ semantic segmentation. Specifically, to overcome the drawback of a limited field of view, we incorporate a densely multi-scale pyramid module and feature fusion module, which can also enhance the global context information. In addition, the system integrates an edge supervision network to refine the segmented results on the decoding side. <b>Results</b>: Experimental results show that DeepVinci can achieve state-of-the-art accuracy, obtaining dice similarity coefficient and mean pixel accuracy values of 0.684 and 0.700, respectively. <b>Conclusions</b>: The proposed DeepVinci network presents a practical and competitive semantic segmentation solution for da Vinci gynecological surgery.https://www.mdpi.com/2075-4418/15/15/1917artificial intelligenceda Vinci Robotgynecological surgeryorgan semantic segmentationdeep learning
spellingShingle Li-An Tseng
Yuan-Chih Tsai
Meng-Yi Bai
Mei-Fang Li
Yi-Liang Lee
Kai-Jo Chiang
Yu-Chi Wang
Jing-Ming Guo
DeepVinci: Organ and Tool Segmentation with Edge Supervision and a Densely Multi-Scale Pyramid Module for Robot-Assisted Surgery
Diagnostics
artificial intelligence
da Vinci Robot
gynecological surgery
organ semantic segmentation
deep learning
title DeepVinci: Organ and Tool Segmentation with Edge Supervision and a Densely Multi-Scale Pyramid Module for Robot-Assisted Surgery
title_full DeepVinci: Organ and Tool Segmentation with Edge Supervision and a Densely Multi-Scale Pyramid Module for Robot-Assisted Surgery
title_fullStr DeepVinci: Organ and Tool Segmentation with Edge Supervision and a Densely Multi-Scale Pyramid Module for Robot-Assisted Surgery
title_full_unstemmed DeepVinci: Organ and Tool Segmentation with Edge Supervision and a Densely Multi-Scale Pyramid Module for Robot-Assisted Surgery
title_short DeepVinci: Organ and Tool Segmentation with Edge Supervision and a Densely Multi-Scale Pyramid Module for Robot-Assisted Surgery
title_sort deepvinci organ and tool segmentation with edge supervision and a densely multi scale pyramid module for robot assisted surgery
topic artificial intelligence
da Vinci Robot
gynecological surgery
organ semantic segmentation
deep learning
url https://www.mdpi.com/2075-4418/15/15/1917
work_keys_str_mv AT liantseng deepvinciorganandtoolsegmentationwithedgesupervisionandadenselymultiscalepyramidmoduleforrobotassistedsurgery
AT yuanchihtsai deepvinciorganandtoolsegmentationwithedgesupervisionandadenselymultiscalepyramidmoduleforrobotassistedsurgery
AT mengyibai deepvinciorganandtoolsegmentationwithedgesupervisionandadenselymultiscalepyramidmoduleforrobotassistedsurgery
AT meifangli deepvinciorganandtoolsegmentationwithedgesupervisionandadenselymultiscalepyramidmoduleforrobotassistedsurgery
AT yilianglee deepvinciorganandtoolsegmentationwithedgesupervisionandadenselymultiscalepyramidmoduleforrobotassistedsurgery
AT kaijochiang deepvinciorganandtoolsegmentationwithedgesupervisionandadenselymultiscalepyramidmoduleforrobotassistedsurgery
AT yuchiwang deepvinciorganandtoolsegmentationwithedgesupervisionandadenselymultiscalepyramidmoduleforrobotassistedsurgery
AT jingmingguo deepvinciorganandtoolsegmentationwithedgesupervisionandadenselymultiscalepyramidmoduleforrobotassistedsurgery