Traditional craniotomy relies on the surgeon's experience and can be complicated owing to excessive skull bone removal, undesirable brain tissue penetration, or severe bleeding. For craniotomy, we developed a robot system based on intraoperative cone-beam computed tomography image guidance and human-robot cooperative interaction, aiming to improve the safety and accuracy of surgery and reduce the labor-intensiveness of the procedure. Intraoperative cone-beam computed tomography image guidance was adopt to improve the accuracy in our experiment. Craniotomy was performed using an interactive method based on human-robot collaboration, which could achieve a natural interactive method in accordance with surgeons' operating habits. The frequency-based method of contact distinction and the method of torque estimation were used to improve the safety of the designed robot. An animal experiment was conducted to verify the effectiveness of the robot system. During the drilling process, the position error was 0.92 ± 0.17 mm (upper surface) and 0.97 ± 0.11 mm (lower surface), and the angle error was 3.37 ± 1.43°. During the milling process, the position error was 1.06 ± 0.13 mm (upper surface) and 1.09 ± 0.09 mm (lower surface). The results showed that the system had sufficient precision and could better complete craniotomy with human-robot collaboration. In addition, with the feedback of multisensor information, the robot system could achieve a sufficient level of safety. The robot system can achieve accurate positioning and safe user-friendly human-robot interaction, which solves problems encountered in the drilling and milling of craniotomy, meets clinical needs, and provides a new method for robot-assisted craniotomy.
Read full abstract