Results of CO2 robotic laser oseotomy in surgery with motion compensation

Abstract

This paper presents a visual servoing application with and without motion compensation and a fixed visual servoing configuration for CO2 laser osteotomy. A multi camera system from ART is used to track the position of the robot and a skull via marker spheres that are attached to both. A CT scan from the skull is performed and segmented to acquire a 3D model. Inside the model the position for the robot for the laser ablation is planned. The accuracy of the lightweight robot is increased with the additional supervision of an optical tracking system. Accuracy improvement was measured with an FARO measurement arm. A visual servoing control schema is presented. The demonstrator shows a working visual servoing application for laser osteotomy. To improve the error resulting mainly from the delay to acquire the data from the devices a motion compensation algorithm is introduced based on iterative learning and a normalized Least Mean Square (nLMS) filter. The results during the simulation and the experimental setup are shown. The system was then evaluated with the CO2 laser system OsteoLas X10 from Caesar - Bonn, Germany. Different cuts are performed with the robot and the CO2 laser system. For the breathing motion a robotic breathing simulator is used. The reached accuracy and the cutting results on bone are shown.