Neurosurgery Resident, PGY-4 University of Pennsylvania Philadelphia, Pennsylvania, United States
Introduction: A core technical competency for spinal surgeons is pedicle instrumentation to ensure neurovascular protection and biomechanical optimization. Augmented reality (AR) is an emerging technology that has the potential to be a useful educational tool for neurosurgical residents and medical students who are not familiar with free-hand (FH) pedicle screw placement by projecting virtual pedicle trajectories using computerized headsets. We quantify the relative assistance of AR by comparing AR and FH pedicle screw accuracy across different levels of surgical experience.
Methods: A spine fellowship-trained board-certified attending neurosurgeon, PGY-4 neurosurgery resident, and 2nd-year medical student placed a total of 32 FH and 32 AR-assisted thoracolumbar pedicle screws in 3 cadavers. A cableless, voice-activated AR system (VisAR, Novarad, Provo, UT) was paired with a HoloLens 2 headset (Microsoft Corp., Redmond, WA). Accuracy was assessed using Chi-square analysis and the Gertzbein-Robins scale (GRS).
Results: The attending had 91.6% (11/12) FH accuracy and 83.3% (10/12) AR accuracy (p=0.54); the resident had 89.9% (8/9) FH accuracy and 100% (9/9) AR accuracy; the medical student had 36.4% (4/11) FH accuracy and 72.7% AR accuracy (8/11), approaching significance (p=0.087). One screw placed by the medical student was GRS D at T8. The medical student had significantly lower FH accuracy when compared to the resident (p=0.017) and attending (p=0.005), but no difference when using AR. FH screw placement was faster by both the attending (median 46s vs. 94.5s, p=0.0047) and resident (median 144s vs. 140s, p=0.05). Total GRS A FH accuracy was 23/32 (71.9%) and 27/32 (84.4%) for AR (p=0.23). The mean AR angle of error between the planned pedicle trajectory and the actual trajectory was 6.2 degrees (95% CI 4.6-7.9)
Conclusion : AR screw placement using a commercially available headset with customized software allowed an inexperienced medical student to place pedicle screws with doubled accuracy compared to his own FH technique in a cadaver. AR technology has the potential to shorten the learning curve necessary to master pedicle screw placement using anatomic landmarks only. With subsequent iterations, this promising technology could be adopted in actual surgical procedures to increase operative efficiency and accuracy of spine instrumentation with very little added cost.
