Spatial Registration for Augmented Reality Spinal Surgery Navigation: A 2D/3D Registration Approach

Objective: The key to spinal surgery navigation based on augmented reality (AR) technology is to integrate reconstructed patient medical images with the patient in the real world space. This study proposes a spatial registration method based on 2D/3D registration to achieve spatial registration between the medical image coordinate system and the patient coordinate system, thereby providing accurate and reliable image guidance for augmented reality spinal surgery navigation.\n\nMethods: Personalized markers that can be detected by both X-ray and monocular camera were designed. Using grayscale-based 2D/3D registration technology, preoperative CT images were registered with intraoperative X-ray images, and the Powell method was used for data optimization. After completing the spatial registration between the medical image coordinate system and the patient, virtual image and real spine fusion were achieved in the augmented reality spinal surgery navigation system. Finally, 2D/3D image registration experiments were conducted, and the surface registration accuracy of the spatial registration method was verified using a spine model.\n\nResults: In the image registration experiment, the success rate of spine model registration was 0.9. The medical image registration of two patients was successful. In the spatial registration experiment, the surface registration errors of the spine model ranged from 0.361 to 0.612 mm, with an overall average surface registration error of 0.501 mm.\n\nConclusion: The spatial registration method based on 2D/3D registration technology can be used in augmented reality spinal surgery navigation systems, and it is highly accurate and minimally invasive.