ABSTRACT
BACKGROUND: Robotic automation and haptic guidance have multiple applications in neurosurgery. OBJECTIVE: To define the spectrum of cranial procedures potentially benefiting from robotic assistance in a university hospital neurosurgical practice setting. METHODS: Procedures utilizing robotic assistance during a 24-mo period were retrospectively analyzed and classified as stereotactic or endoscopic based on the mode utilized in the ROSA system (Zimmer Biomet, Warsaw, Indiana). Machine log file data were retrospectively analyzed to compare registration accuracy using 3 different methods: (1) facial laser scanning, (2) bone fiduciary, or (3) skin fiduciary. RESULTS: Two hundred seven cranial neurosurgical procedures utilizing robotic assistance were performed in a 24-mo period. One hundred forty-five procedures utilizing the stereotactic mode included 33% stereotactic biopsy, 31% Stereo-EEG electrode insertion, 20% cranial navigation, 7% stereotactic catheter placement, 6% craniofacial stereotactic wire placement, 2% deep brain stimulation lead placement, and 1% stereotactic radiofrequency ablation. Sixty-two procedures utilizing the haptic endoscope guidance mode consisted of 48% transnasal endoscopic, 29% ventriculoscopic, and 23% endoport tubular access. Statistically significant differences in registration accuracies were observed with 0.521 ± 0.135 mm (n = 132) for facial laser scanning, 1.026 ± 0.398 mm for bone fiduciary (n = 22), and 1.750 ± 0.967 mm for skin fiduciary (n = 30; ANOVA, P < .001). CONCLUSION: The combination of accurate, automated stereotaxy with image and haptic guidance can be applied to a wide range of cranial neurosurgical procedures. The facial laser scanning method offered the best registration accuracy for the ROSA system based on our retrospective analysis.
