Mixed reality enhances interobserver reliability in tibial plateau fracture classification compared to computed tomography, 3D computed tomography and 3D printing

Abstract

Purpose: Tibial plateau fractures present a diagnostic challenge due to complex morphology and demanding anatomy. Reliable diagnosis is essential for treatment planning and clinical outcomes. Prior studies showed benefits of 3D computed tomography (3DCT) and 3D printing (3DP) using CT-based systems, but the potential of mixed reality (MR) remains unexplored. This study tested the hypothesis that MR enhances spatial understanding of fracture morphology, thereby improving interobserver reliability and diagnostic confidence across modern classification systems compared to CT, 3DCT and 3DP. Methods: Twenty-two intra-articular tibial plateau fractures (Arbeitsgemeinschaft Osteosynthese [AO]/OTA type B or C) were assessed using four visualisation modalities: (1) CT, (2) 3D reconstructions, (3) 3D-printed models and (4) MR holograms (Microsoft HoloLens 2). Twelve raters (six junior and six senior surgeons) classified each fracture using AO/OTA, 10-segment and revisited Schatzker systems and rated their diagnostic confidence using a 5-point Likert scale. Interobserver reliability was measured via kappa coefficients and percentage match. Differences were analysed using one-way analysis of variance (p < 0.05). Results: MR demonstrated the highest interobserver reliability for all classification systems: AO/OTA (0.43 vs. 0.32 for CT), 10-segment (0.17 vs. 0.11) and revisited Schatzker (0.43 vs. 0.24). Diagnostic confidence for AO/OTA differed significantly between modalities (p = 0.045), with the highest values observed for 3DP (76%) and MR (58%) compared to CT (47%). Revisited Schatzker showed the largest reliability gain, though confidence declined compared to 3DP (59% vs. 34%). Conclusion: MR improved classification reliability and diagnostic confidence for tibial plateau fractures compared to CT, 3DCT and 3DP. The effect was most evident in less-experienced surgeons, indicating MR's value for surgical training and preoperative planning. Further research should explore MR's integration into education and its impact on clinical outcomes. Level of Evidence: Level II.