Percutaneous Pedicle Screw Placement in the Cervical Vertebrae Using Augmented Reality-assisted : A Cadaveric Proof-of-concept Study

Introduction: Augmented reality-mediated spine surgery (ARMSS) can improve the effectiveness, precision, and safety of traditional percutaneous pedicle screw placements. Previous data from cadaveric studies and in real patients demonstrated the clinical utility of ARMSS for guiding instrumentation of the thoracolumbar spine. However, the technical feasibility of ARMSS for the cervical spine is unknown. The objective of this cadaveric study was to assess the accuracy and precision of augmented reality (AR) guided percutaneous pedicle implant insertion in the cervical spine.

Methods: AR-guided instrumentation was performed in 5 cadaveric torsos at levels ranging from C3 to C7. Postprocedural CT scans were graded for clinical accuracy by two independent neuroradiologists using the Gertzbein-Robbins scale. Technical precision was calculated via superimposition analysis using the Medical Image Interaction Toolkit to yield angular trajectory (°) and linear screw tip (mm) deviation from virtual pedicle screw position, as recorded by the ARMSS platform, by actual pedicle screw position, derived from the postprocedural imaging.

Results: In total, 41 cervical screws (C3-C7, Pedicle Screws and Lateral Mass) were positioned in 5 cadavers. Each surgeon (n=5) navigated at least 7 screws. Clinical accuracy (Gertzbein-Robbins grade A or B) was found to be 100%. The mean values of tip and trajectory errors were less than 3mm and 30°, respectively (p < 0.05). No instrumented cervical screw had cortex violation over 2mm. The 99% upper bound limits for tip and trajectory error were 2.42mm and 2.94mm, respectively.

Conclusion : These accuracy findings are similar to literature values for cervical screw placement accuracy using image-guided navigation systems. This demonstrates that AR-guided cervical pedicle implant insertion is technically feasible, accurate, and highly precise. AR-guided percutaneous screw placement at C3-C7 levels produces accuracy results comparable to those reported for instrumentation of the thoracolumbar spine.