Related Experiment Video
Updated: Jun 14, 2025
07:46
Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility
Published on: August 9, 2024
673
Automated surgical skill assessment in colorectal surgery using a deep learning-based surgical phase recognition
Kei Nakajima1,2, Daichi Kitaguchi1, Shin Takenaka1
1Department for the Promotion of Medical Device Innovation, National Cancer Center Hospital East, 6-5-1 Kashiwanoha, Kashiwa, Chiba, 277-8577, Japan.
Surgical Endoscopy
|August 30, 2024
Summary
Automated surgical skill assessment is feasible using a deep learning model that recognizes surgical phases. This AI model accurately distinguished between expert, intermediate, and novice surgeons based on performance metrics.
Area of Science:
- Minimally Invasive Surgery
- Surgical Education
- Artificial Intelligence in Medicine
Background:
- Addressing the need for objective surgical skill assessment to overcome manual evaluation limitations.
- Exploring automated methods to reduce subjectivity and bias in surgical training.
Purpose of the Study:
- To verify the feasibility of a surgical phase recognition model for automated surgical skill assessment.
- To evaluate the model's ability to differentiate skill levels in laparoscopic sigmoidectomy.
Main Methods:
- Developed a deep learning model to recognize five surgical phases of laparoscopic sigmoidectomy.
- Assessed model performance in distinguishing between expert, intermediate, and novice skill groups based on Endoscopic Surgical Skill Qualification System (ESSQS) scores and surgical experience.
- Utilized multiple regression analysis to evaluate the correlation between model-derived parameters and ESSQS scores across 1,272 videos.
Main Results:
- Expert surgeons demonstrated significantly shorter times for colon mobilization, mesorectal dissection, rectal transection, and anastomosis compared to intermediate and novice groups.
- Phase transition counts were significantly lower in expert surgeons.
- Higher Endoscopic Surgical Skill Qualification System (ESSQS) scores correlated with higher AI-derived scores, indicating model accuracy.
Conclusions:
- The developed surgical phase recognition model shows potential for automated surgical skill assessment.
- This AI-driven approach can provide objective feedback for surgical training and evaluation.
