Research On The Laparoscopic Vision-based Navigation Method For Robot Assisted Surgery

Robot assisted surgery is a new type of surgery developed in recent 30 years, which has advantages of little trauma and faster recovery. But surgical robots cannot instead of surgeon,the main role of it is to assist the surgeon during operation and almost no autonomy especially for the robot arm which clamps laparoscope. During laparoscopic surgery, the surgeon currently must instruct a human camera operator or a robotic arm to move the camera. This process is inefficiency, and the camera is not always placed in an ideal position. To solve these problems, we have developed a laparoscopic vision-based navigation method that tracks laparoscopic instruments.For minimally invasive laparoscopic surgery robot system understudied in our laboratory,the motion-image model of laparoscope is established combined with the imaging model and robotic kinematics, which lay the ground for laparoscopic vision-based navigation. Combined with the environment and the process of robot assisted surgery, the key technology of associated analysis of laparoscopic vision-based navigation is analyzed and the overall process of laparoscopic vision-based navigation is presented.According to the commonly used in the process of surgical operation, operation status is set. The region of interest (output image) during the navigation is defined combined with the operation status. The laparoscopic movement decision method is developed based on the region of interest, which is used to determine when and how the laparoscope move. And the calculation process of the relevant parameters is presented detailly.Image segmentation algorithm based on mean shift and tracking-by-detection with kernels is analyzed. Combined with the demand of laparoscopic vision-based navigation,these two kinds of algorithm is improved respectively which make it suitable for laparoscopic vision-based navigation method. Finally, the improved algorithm and the original algorithm are analyzed through experiments, and the advantages of the improved algorithm are proved.Finally, the program of laparoscopic image processing is write on VS2013 using a variety of external dependencies.The positioning method of surgical instruments based on machine vison is analyzed in detail and the experimental environment is established. The experiments of camera calibration and accurate and quick positioning are implemented. Combined with the results of surgical instruments location experiment, the laparoscopic vision-based navigation method is dynamically simulated using MATLAB. The feasibility of this method is proved through the laparoscopic movement condition and the change curve of the parameters of output image during the simulation.