Real Time Video Based Detection Of Retained Surgical Instruments During Intraoperative Procedures

Retained surgical instruments post-surgery remain a significant threat to patient safety, necessitating novel solutions. In response to the challenges of obscured instruments in blood-soaked environments and under tissue layers, this research introduces a comprehensive approach. Leveraging real-time surgical video acquisition, advanced image pre-processing, and the You Only Look Once (YOLO) algorithm, our system enables the swift and accurate detection and tracking of surgical instruments during procedures. The proposed methodology involves real-time video acquisition, followed by image enhancement techniques to address challenging surgical conditions. The YOLO algorithm is then implemented for efficient object detection and tracking. This approach ensures the continuous monitoring of specific instruments, significantly reducing the risk of retained surgical objects. An additional algorithm is introduced for the instrument verification, providing a warning, and pinpointing the location of any mismatched object. This research represents a promising step towards improving patient safety during surgeries by employing cutting-edge technology to address the complex issue of retained surgical instruments. The real-time tracking capabilities and accuracy of the YOLO algorithm contribute to the reliability and effectiveness of the proposed system, offering a potential breakthrough in reducing adverse events associated with surgical procedures.