Related Experiment Video
Updated: Nov 23, 2025
06:39
Three-dimensional Location Approach with Silk Thread Guided Laparoscopic Segmentectomy for Liver Tumor
Published on: May 23, 2025
257
Deep learning visual analysis in laparoscopic surgery: a systematic review and diagnostic test accuracy meta-analysis
Roi Anteby1,2, Nir Horesh3,4, Shelly Soffer3
1Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel. roianteby@mail.tau.ac.il.
Surgical Endoscopy
|January 5, 2021
Summary
Deep learning shows promise for analyzing laparoscopic surgery videos, with high accuracy in recognizing surgical elements. However, methodological limitations and bias risk highlight the need for clinician involvement to advance artificial intelligence in surgery.
Area of Science:
- Artificial Intelligence in Medicine
- Surgical Technology
- Medical Image Analysis
Background:
- Deep learning (DL) has transformed medical image processing and shows potential for advancing laparoscopic surgery.
- The study evaluates the accuracy of DL networks in analyzing laparoscopic surgery videos.
Purpose of the Study:
- To assess the diagnostic performance of deep learning models in analyzing videos of laparoscopic procedures.
- To identify common applications and surgical procedures analyzed by DL in laparoscopy.
Main Methods:
- A systematic review of Medline, Embase, IEEE Xplore, and Web of Science databases (2012-2020).
- Inclusion of studies testing deep learning models, specifically convolutional neural networks, for laparoscopic video analysis.
- Meta-analysis using a random effects model to estimate pooled sensitivity and specificity.
Main Results:
- 32 studies met inclusion criteria, analyzing 3004 videos, primarily for instrument and phase recognition in cholecystectomy and gynecological surgeries.
- Pooled sensitivity was 0.93 (95% CI 0.85-0.97) and specificity was 0.96 (95% CI 0.84-0.99) in studies with sufficient data.
- The majority of included studies exhibited a high risk of bias.
Conclusions:
- Deep learning research in laparoscopic surgery demonstrates significant potential but faces methodological limitations.
- Clinician collaboration is crucial for advancing artificial intelligence in surgery through standardized data and reporting.
