Towards Seamless Surgical Guidance: A Robust Marker-Based Multi-camera AR Navigation System with Advanced Calibration and Detection Techniques

In recent years, many marker-based augmented reality (AR) surgical navigation systems have been developed to assist with surgeries. However, the current marker-based registration methods encounter challenges related to complex marker placement and adjustments, limiting their widespread adoption in clinical settings, especially in diverse surgical scenarios. This study presents a novel marker-based multi-camera AR surgical navigation system that encompasses a robust marker detection framework and a high-precision registration system. Our marker-based registration system addresses the challenges of marker placement by utilizing an extended perception range and a high-accuracy calibration solution. Moreover, we introduce an innovative deep learning-based framework designed to provide accurate and robust marker detection even under challenging lighting and motion conditions. Validation experiments underscore the superiority of our system over standard methods, exhibiting enhanced performance in terms of marker detection, marker identification, and calibration accuracy. Our proposed system demonstrates high registration accuracy in real-life scenarios, highlighting its potential and suitability in clinical settings.