Related Experiment Video
Updated: May 6, 2026
05:57
A Teleoperated Robotic System-Assisted Percutaneous Transiliac-Transsacral Screw Fixation Technique
Published on: January 6, 2023
3.2K
Enhancing intraoperative safety through an intelligent surgical instrument verification system based on computer
Tsan-Wen Huang1,2, Jing Ting Wong3, Ming-Chuan Chiu3
1Department of Orthopaedic Surgery, Jen-Ai Hospital, Taichung, 412224, Taiwan.
Summary
This study introduces an intelligent system for surgical instrument counting, enhancing patient safety by integrating visual detection and weight verification. The prototype demonstrates high accuracy and efficiency, paving the way for safer perioperative management.
Area of Science:
- Medical Technology
- Computer Vision
- Artificial Intelligence
Background:
- Manual surgical instrument counting is time-consuming (15-30 minutes) and error-prone in high-pressure operating room environments.
- Increasing instrument complexity heightens identification challenges and the risk of omissions during surgery.
- Current methods pose risks to patient safety and operational efficiency in perioperative care.
Purpose of the Study:
- To design, integrate, and validate a multimodal verification framework for high-risk orthopedic surgical instrument counting.
- To enhance patient safety and operational efficiency through real-time instrument tracking and discrepancy checking.
- To develop an intelligent system addressing the limitations of traditional manual counting methods.
Main Methods:
- Developed a system with four integrated modules: image acquisition/annotation, YOLOv11 for instrument detection, PaddleOCR for scale reading extraction, and an intelligent alert system.
- Focused on systematic design, integration, and technical validation of a multimodal framework for orthopedic instrument counting.
- Employed YOLOv11 for precise instrument identification and PaddleOCR for accurate digital extraction of scale readings.
Main Results:
- YOLOv11 achieved high performance in instrument identification: 0.91 precision, 0.88 recall, 0.89 F1-score, and 0.93 mAP50.
- PaddleOCR demonstrated 99.7% accuracy in digit recognition with a response time under 0.5 seconds for near-real-time extraction.
- The system showed strong technical feasibility and potential clinical relevance for perioperative instrument management.
Conclusions:
- An intelligent surgical instrument verification prototype was established, proving the feasibility of a multimodal safety-checking framework.
- Key contributions include a prototype integrating detection, scale reading extraction, and alerting, complemented by weight-based cross-verification.
- The modular system architecture offers potential for extension to broader perioperative safety management, pending further usability and workflow compatibility studies.
