Pedicle Screw Placement With Augmented Reality Versus Robotic-Assisted Surgery.

Abstract

This was a single-center prospective clinical and radiographic analysis of pedicle screw instrumentation with robotic-assisted navigation (RAN) and augmented reality (AR). This study aimed to compare the accuracy of lumbosacral pedicle screw placement with RAN versus AR. RAN and AR have demonstrated superior accuracy in lumbar pedicle screw placement compared with conventional free-hand techniques. RAN and AR techniques both tout-specific advantages over their counterparts, but to date, no study has directly compared the two technologies regarding pedicle screw accuracy. Patients who underwent RAN or head-mounted AR navigated lumbosacral (L1-S1) pedicle screw placement for degenerative conditions were included. Screw accuracy was assessed by two independent reviewers on intraoperative 3D fluoroscopic scans using the Gertzbein and Robbins scale. A generalized linear mixed model was applied to evaluate the relationship between the screw placement technique and accuracy. Two hundred and twelve patients undergoing lumbosacral instrumentation with a total of 1211 pedicle screws placed using RAN (n=108; screws=827) or AR (n=104; screws=384). Overall, grade A was achieved in 92.6% of screws. No significant difference was found between RAN and AR screw placement regarding the incidence of accurate (grade A and B screws; RAN n=824; 99.6% and AR n=379, 98.7%) versus inaccurate screws (grade C and D screws; RAN n=3, 0.4% and AR n=5, 1.3%). When comparing "optimal" grade A screws (RAN n= 787, 95.2%, AR n=345, 89.8%) versus all other screws (B, C, and D), significantly higher accuracy was achieved using RAN ( P =0.001). RAN and AR both achieved high accuracy in lumbosacral pedicle screw placement, proving reliable for this procedure. However, RAN resulted in significantly more grade A screw placements than AR.