Validation of Augmented Reality-Guided Implant Positioning in Cadaveric Models: A Step Toward Clinical Application

Introduction: Accurate placement of patient-specific implants is essential in orthopaedic oncology, particularly in complex pelvic resections.Augmented reality (AR)-assisted navigation has shown promise in experimental settings, yet its validation in anatomically realistic environments remains limited.This study assesses, for the first time, the precision of a novel ARbased surgical guidance system in human cadaveric specimens for the placement of custom acetabular implants.Materials and Methods: Ten cadaveric hemipelves were used.Preoperative planning was performed using CT imaging and 3D segmentation to design patient-specific implants.A Microsoft HoloLens 2 headset was employed alongside a custom-developed AR application, enabling holographic visualisation of the planned implant over the surgical field.Intraoperative guidance was provided via AR markers and a real-time colour-coded feedback system.Postoperative CT scans were acquired and compared to preoperative plans through 3D registration to quantify positioning accuracy.Results: The mean translational error was 3.79 mm (SD 2.09 mm) and the mean angular error was 3.73 (SD 2.59), both within clinically acceptable thresholds.Compared to previous phantom-