Vision-based object detection and tracking from surgical microscope video for augmented cochlear implant surgery

Key Points

• The aim is to enhance object detection and tracking during cochlear implant surgery using video from surgical microscopes.
• Integrated YOLOv9 for object detection
• Utilized BoT-SORT for tracking
• Evaluated identity preservation and trajectory consistency
• Maintained stable identity preservation during procedures
• Achieved trajectory consistency for monitored objects
• Demonstrated strong potential for real-world surgical applications