Deep Learning Applications in Evaluating and Predicting Technical Surgical Skills in Robotic-Assisted Vaginal Cuff Closure: A Multicenter Prospective Study

Highlights

• Deep learning techniques can objectively assess surgical skills and detect technical errors during robotic-assisted vaginal cuff closure, supporting surgical education.
• Multimodal learning models integrating video data achieve high accuracy (>80%) in skill assessment, correlating strongly with validated human expert ratings.
• Objective metrics such as Modifiable Global Evaluative Assessment of Robotic Skills (GEARS) and Objective Clinical Human Reliability Analysis (OCHRA) correlate with surgeon experience and operative outcomes.
• These methods lay foundational work toward AI-driven quality monitoring and evidence-based credentialing in minimally invasive gynecologic surgery.