Feasibility of Camera Calibration and Surface Registration Using HoloLens 2 Built-in Cameras and Point Cloud Scanning for Surgical Navigation

This study evaluated the feasibility of camera calibration and surface point cloud registration using the built-in cameras of HoloLens 2 (HL-2) for surgical navigation. Intrinsic and extrinsic parameters of single or multiple built-in cameras were assessed using the checkerboard method and compared with those obtained using alternative methods. Additionally, a 3D-printed head phantom was scanned using HL-2’s depth, environment, and PV cameras, and surface registration was executed using the Iterative Closest Point (ICP) algorithm. The results indicated high calibration reliability, satisfactory point cloud extraction, and successful surface registration, highlighting factors that affect accuracy and efficiency.