Accuracy of a novel mixed reality surgical platform for total knee arthroplasty.

BACKGROUND: Computer-assisted navigation has improved surgeons' ability to achieve accurate implant placement in total knee arthroplasty (TKA). As technology evolves and new systems are introduced, it is imperative to evaluate their accuracy for achieving desired goals. METHODS: This work evaluated the accuracy of a novel mixed reality surgical guidance platform that uses a head-mounted device to measure patient bony anatomy, quantify soft tissue balance, and provide quantitative resection guidance overlaid on the surgical field. Accuracy is evaluated in a cadaveric simulated use study and in a comprehensive evaluation of the tracking subsystem using the international ASTM standard F2554-22. Depth and angular errors are reported across eight knees for the proximal tibial, distal femoral, and posterior femoral resections by comparing platform-navigated resection metrics to caliper-measured resection depths and post-operative CT-measured angles. Analysis of the tracking subsystem investigated instrument localization error across the tracking volume, including rotational and positional extremes. Testing also extended the ASTM standard to include additional rotation tests and an evaluation procedure for planar accuracy metrics. RESULTS: All cadaveric simulated use depth and angular absolute errors were below 2 mm and 2°, with 83% less than or equal to 1 mm and 1°. Absolute resection errors averaged 0.7 ± 0.4 mm and 0.6 ± 0.4° for depths and angles, respectively. The tracking subsystem localized over 99.5% of points with positional accuracy better than ± 2 mm and localized over 99.5% of planes with angular accuracy better than ± 1°. Average absolute tracking errors were sub-millimeter and sub-degree. CONCLUSION: This open and imageless platform for TKA surgical guidance requires only 13 landmarks with no additional equipment footprint, minimal tooling, and data overlaid holographically onto the surgical scene, while still providing a comprehensive set of metrics with state-of-the-art accuracy. Video Abstract.