Clinical Report: AI-Driven Recognition and Assessment of Surgical Phases in SILC
Overview
This study developed a deep learning model for automatic surgical phase recognition in single-incision laparoscopic cholecystectomy (SILC) videos, assessing its accuracy.
Background
Surgical process modeling (SPM) enhances the understanding of surgical workflows, which is crucial for improving surgical outcomes. The integration of deep learning technologies in surgical phase recognition holds promise for real-time decision-making and skill assessment.
Data Highlights
Dataset
Number of Videos
Image Samples
SILC Videos
148
127,496
Key Findings
Deep learning model developed for automatic recognition of surgical phases in SILC.
Model trained on 148 SILC videos, including 20 non-insufflation cases.
Videos were annotated into six distinct surgical phases.
Each video was sampled at 1-second intervals to create a comprehensive dataset.
Study approved by the ethics committee of the Second Affiliated Hospital Zhejiang University School of Medicine.
Clinical Implications
The development of an AI model for surgical phase recognition in SILC could enhance the quality of surgical education and training. Accurate phase identification may also support real-time decision-making during procedures.
Conclusion
The study demonstrates the feasibility of using deep learning for surgical phase recognition in SILC, paving the way for future advancements in surgical process modeling.
