Related Experiment Video
Updated: Feb 20, 2026
07:22
Surgical Robot-Assisted Transanal Specimen Extraction Radical Sigmoidectomy Without an Auxiliary Abdominal Incision
Published on: June 13, 2025
1.0K
Surgical-tools detection based on Convolutional Neural Network in laparoscopic robot-assisted surgery.
Summary
This study introduces real-time Convolutional Neural Network (CNN) models to detect surgical instruments during laparoscopic surgery, aiming to prevent damage caused by limited visibility and tactile feedback in robotic systems. The models achieved 72.26% mean average precision.
Area of Science:
- Medical technology
- Computer vision
- Surgical robotics
Background:
- Laparoscopic surgery offers faster recovery and less pain but faces challenges.
- Robotic systems in laparoscopy can cause instrument, organ, or tissue damage.
- Limitations include narrow fields of view, confined operating spaces, and poor tactile feedback.
Purpose of the Study:
- To develop real-time Convolutional Neural Network (CNN) models for surgical instrument detection.
- To enhance safety and precision in laparoscopic surgery by mitigating risks associated with robotic systems.
Main Methods:
- Utilized a dataset comprising 7 surgical tools for CNN model training.
- Applied the unified architecture of YOLO (You Only Look Once) for real-time surgical instrument tracking.
- Evaluated model performance using precision and recall metrics.
Main Results:
- Achieved a mean average precision of 72.26% across the dataset.
- Demonstrated the feasibility of using CNNs and YOLO for real-time instrument detection in laparoscopy.
Conclusions:
- The proposed CNN models show promise in improving the safety of laparoscopic surgery.
- Real-time instrument detection can help prevent unintended damage during minimally invasive procedures.
