An automated robotic approach with redundant navigation for minimal invasive extended transsphenoidal skull base surgery

Abstract

Background: The aim of this work was to determine the feasibility of a robotic-assisted and fully automated approach to the sphenoid sinus. An image-guided robotic system was designed to address potential human errors in performing transsphenoidal sinus surgery by combining the reproducible accuracy of a robotic system with standard computer navigation.

Methods: A six-degrees of freedom robotic assistance system and an opto-electrical navigation system were combined for image-guided assistance with redundantly controlled robotics. Newly designed endoscopic instruments for robotic surgery have been developed and are described. Telemanipulatory, as well as fully automated procedures, were tested on cadaveric heads as part of a preclinical trial.

Results: A fully automated sphenoidotomy as well as a telemanipulatory sphenoidectomy were performed successfully on cadaveric heads. Intraoperative performance, accuracy assessment studies, as well as possible sources of stereotactic offsets are described. The mean measured robotic reproducibility accuracy was 0.056 mm (range: 0.02 - 0.14 mm) and the mean overall navigated robotic accuracy, including all transformation and registration errors was 1.53 mm (range: 1.13 - 1.89 mm) respectively.

Conclusion: A system for robot-guided surgery in combination with redundant navigational control was developed. It allows highly accurate maneuvers, performed either in a telemanipulation mode as master-slave system or in a fully automated fashion. A sphenoidectomy on cadaveric heads was performed in both telemanipulation and fully automated modes. The overall intraoperative accuracy was in the range of the resolution of the CT images and stereotactic offsets were caused mainly due to deflections of the endoscopic operating instrument.