Early Clinical Safety Profile and Accuracy of Augmented Reality-Assisted Pedicle Screw Placement: A Multi-Center...

Retrospective case series.

To report on the early clinical safety profile and accuracy of Augmented reality (AR)-assisted pedicle screw placement in patients undergoing minimally invasive spinal surgery.

While navigation-guided pedicle screw placement has improved accuracy rates to 89-100%, current systems face challenges, including line-of-sight interruptions and attention shift from patient to remote screens. AR technology projects navigation data directly onto the surgical field, potentially overcoming these limitations while maintaining high accuracy.

A multi-center retrospective analysis was conducted across two institutions, reviewing 31 procedures involving 200 pedicle screws placed using the Xvision AR navigation system (Augmedics Inc.). All procedures were performed by fellowship-trained spine surgeons. Accuracy was assessed using pre- and post-instrumentation CT O-arm imaging according to the Gertzbein and Robbins classification system. Novel angular deviation analysis was performed to evaluate trajectory accuracy. Clinical outcomes, including complications and revision rates, were recorded.

Among 200 pedicle screws analyzed, 84.5% were classified as Gertzbein-Robbins grade A (complete containment), 13.5% as grade B (trivial breach), and 2% as grade C, yielding an overall accuracy rate of 98%. The pedicle breach rate was 15.5%, with no instances of superior endplate or facet joint violations. Mean angular deviation from the ideal trajectory was 0.78±4.56 degrees. No screw revisions, hardware-related complications, or symptomatic nerve encroachment occurred. Two patients experienced non-hardware-related systemic infections.

AR-assisted navigation demonstrates high accuracy for pedicle screw placement with minimal angular deviation and low complication rates. This technology offers promising potential for improving surgical precision while maintaining workflow efficiency.

Level IV.