Intraoperative evaluation of bone resection accuracy in total knee arthroplasty using an augmented reality-based...

Computer-assisted surgery (CAS) in total knee arthroplasty (TKA) has seen growing adoption, with over 30% of primary TKA procedures in Japan using CAS as of 2021. Augmented reality (AR) navigation has emerged as a novel alternative, providing real-time visualization and high-precision guidance during surgery. However, standardized protocols, long-term outcomes, and quantitative accuracy data remain limited. This study aimed to evaluate the intraoperative bone resection accuracy achieved with an AR-based navigation system during TKA.

We retrospectively analyzed 100 knees that underwent primary TKA using the NextAR navigation system (Medacta International SA, Switzerland). Bone resection accuracy was assessed by comparing intraoperative cut angles and thicknesses to preoperative plans. Evaluated parameters included femoral coronal, sagittal, and rotational angles; distal and posterior resection thicknesses (medial/lateral); tibial coronal, posterior slope, and rotational angles; tibial resection thicknesses (medial/lateral); and implant size concordance.

Mean absolute differences from preoperative plans were as follows: femoral coronal angle 0.46° ± 0.48°, sagittal angle 0.72° ± 0.61°, and rotational angle 0.30° ± 0.33°; distal cut thickness 0.56 ± 0.46 mm (medial) and 0.54 ± 0.49 mm (lateral); posterior condyle cut thickness 0.23 ± 0.35 mm (medial) and 0.35 ± 0.44 mm (lateral). Tibial parameters showed coronal error of 0.64° ± 0.56°, posterior slope 0.66° ± 0.55°, and rotation 0.25° ± 0.60°; resection thickness 0.73 ± 0.69 mm (medial) and 0.79 ± 0.75 mm (lateral). Implant size matched 100% with preoperative planning.

The AR-based navigation system enabled precise three-dimensional intraoperative assessment and execution of bone resections in TKA. All angular and thickness errors remained within 1° and 1 mm, respectively, confirming high reproducibility and precision across multiple planes.