A Robotic System Integrated With CBCT for Cochlear Implant Surgery: Accuracy Improvement and Validation

In robot-assisted cochlear implantation (CI), the drilling accuracy of a tunnel for electrode array (EA) insertion is critical to the success of the surgery. Currently, robots for tunnel drilling in CI inevitably introduce additional invasiveness, and struggle to ensure consistent accuracy in the drilling process. Based on this application background, a novel configuration of integrating the cone-beam computed tomography (CBCT) into a robotic system to perform CI is proposed. Benefiting from the novel configuration, the registration process is simplified, and high-accuracy drilling can be achieved without additional invasiveness. An error model to calibrate the manipulator kinematic parameters is established through dimension-increased least squares. Calibration experiments show that the accuracy of the manipulator is 0.05 $\pm$ 0.01 mm, which is improved by approximately 25%. A novel offline image-to-robot registration method through matching the point cloud of marker holes drilled on a phantom is proposed. Thirty experiments on temporal bone specimens show that the simulated surgical accuracy of the robotic system is 0.12 $\pm$ 0.07 mm utilizing the proposed calibration and registration methods. Eight experiments on cadaver specimens preliminarily validate the effectiveness of the robotic system.