Related Experiment Video
Updated: Jul 9, 2025
07:27
Robotics in Surgery: A Modular Robotic Platform Driven Gastric Wedge Resection
Published on: February 7, 2025
475
Automatic surgical phase recognition-based skill assessment in laparoscopic distal gastrectomy using multicenter
Masaru Komatsu1,2,3, Daichi Kitaguchi2, Masahiro Yura1
1Gastric Surgery Division, National Cancer Center Hospital East, 6-5-1 Kashiwanoha, Kashiwa, Chiba, 277-8577, Japan.
Summary
This study developed a deep learning model for surgical phase recognition in laparoscopic distal gastrectomy. The model accurately identifies surgical phases and shows potential for automatic surgical skill assessment.
Area of Science:
- Medical Artificial Intelligence
- Surgical Technology
- Computational Pathology
Background:
- Gastric surgery phases can be clearly defined, enabling stylization through deep learning.
- Deep learning-based surgical phase recognition aids in automatic surgical skill assessment.
Purpose of the Study:
- To develop a deep learning model for surgical phase recognition in laparoscopic distal gastrectomy.
- To assess the feasibility of using this model for automatic surgical skill assessment.
Main Methods:
- Utilized multicenter videos of laparoscopic distal gastrectomy from 20 hospitals.
- Developed a deep learning image classification model for nine defined surgical phases.
- Correlated model outputs (frame counts, surgical field adequacy) with manual skill assessment scores.
Main Results:
- Achieved 88.8% overall accuracy in surgical phase recognition.
- High-skill groups showed significantly fewer frames in lymphadenectomy and reconstruction phases (P < 0.01).
- Model-assessed surgical field adequacy was significantly higher in high-skill groups (P = 0.04).
Conclusions:
- The developed deep learning model demonstrates high accuracy in recognizing surgical phases.
- The model holds significant potential for application in automated surgical skill assessment systems.
