Related Experiment Video
Updated: Aug 7, 2025
05:12
Robotized Testing of Camera Positions to Determine Ideal Configuration for Stereo 3D Visualization of Open-Heart Surgery
Published on: August 12, 2021
2.1K
Study of visual SLAM methods in minimally invasive surgery
Liwei Deng1, Zhen Liu1, Tao Zhang1
1Heilongjiang Provincial Key Laboratory of Complex Intelligent System and Integration, School of Automation, Harbin University of Science and Technology, Harbin 150080, China.
Mathematical Biosciences and Engineering : MBE
|March 10, 2023
Summary
This study introduces a visual simultaneous localization and mapping (SLAM) technique for minimally invasive surgery. It enhances endoscope localization and 3D reconstruction, improving surgical visualization and accuracy.
Area of Science:
- Medical Technology
- Computer Vision
- Surgical Innovation
Background:
- Minimally invasive surgery offers benefits but faces limitations like poor depth perception and localization.
- Current endoscopic visualization struggles with 3D information and complete cavity views.
Purpose of the Study:
- To develop a visual simultaneous localization and mapping (SLAM) approach for enhanced endoscope localization and surgical region reconstruction.
- To overcome the limitations of 2D imaging in minimally invasive procedures.
Main Methods:
- Utilized K-Means and Super point algorithms for feature extraction in endoscopic images.
- Employed the iterative closest point (ICP) method for endoscope pose estimation.
- Applied stereo matching to generate disparity maps and reconstruct 3D point clouds.
Main Results:
- K-Means + Super point improved feature matching by 32.69% and reduced error rates.
- The SLAM approach successfully localized the endoscope and reconstructed the surgical area.
- Achieved a 25.28% increase in effective points and a 1.98% decrease in extraction time.
Conclusions:
- Visual SLAM offers a robust solution for endoscope localization and 3D reconstruction in minimally invasive surgery.
- This technique addresses key limitations of current endoscopic visualization, potentially improving surgical outcomes.
