Design Of Robot Assisted Knee Surgery System And Algorithms For Post-surgery Evaluation

With rapid developments of computer science and robotics, surgery operations have entered into a biological intelligence times integrated by biology, informatics and physics. In recent decades, a lot of practical computer assisted surgery (CAS) systems have emerged. Orthopedic surgery is the field that many new technologies and new equipments used in, and computer assisted orthopedic surgery system (CAOS) has become one of the most active branches of CAS.Every year, thousands of patients suffer from joint diseases, such as rheumatoid arthritis or osteoarthritis, and need total knee replacement (TKR) surgery to recover their normal functions. Presently, the positioning of prosthetic components in surgeries mainly depends on clinic experiences of doctors and special surgical guiding devices. To avoid the limitations of jig-based TKR systems, researchers devoted to develop robot/computer assisted surgery systems, with the aid of which a better operation precision and surgical quality are well expected.Based on current robot assisted TKR systems, we designed a visual navigated hand-eye robot assisted TKR surgery system– WATO-II. In WATO-II, both the infrared stereo cameras and the cutting tool are fixed on the end-effecter of the robot. During the surgery, WATO-II automatically determines the position of cutting planes and establishes the mechanical axis according to the infrared markers, then guides the robot complete the surgery. Lots of experiments on gesso bones and human bones demonstrated the high accuracy and efficiency of WATO-II, which laid a good foundation of clinical applications.In CAOS, besides system design, another important aspect is the research of post-surgery evaluation algorithm. Post-knee-surgery evaluation consists of quantitatively evaluation of soft tissue balance and quantitatively evaluation of knee motion mode. Currently, such evaluations are mostly dependent on the experiences of the surgeon and the evaluation results are not so accurate. In order to obtain a reasonable and accurate post-knee-surgery evaluation result, after taking into account that the soft tissue is a typical three-dimensional deformable structure and the knee joint is a typical three-dimensional articulated-rigid structure, we propose a series of computer vision based analysis methods for deformable and articulated-rigid objects, which gives a new way for post-knee-surgery quantitative evaluation.The major contributions of this thesis are as follows:1) Design and realize a pre-clinical WATO-II experimental system. In the system architecture, we design the system components of WATO-II, as well as the structures of the probe, the tag and the cutting tool. In the core algorithms, we design the system calibration module, the tool calibration module and the surgical planning and navigation module. We propose the integration calibration algorithm for the camera, the stereo rig and the robot hand-eye, the infrared probe calibration algorithm, the tool calibration algorithm and physiological point based femoral and tibia positioning and calculation algorithm. WATO-II makes use of the physiological characteristics of the knee and the stereo navigation system, can get high-accurate locations of the femur and the tibia and achieve high-precision cutting of the bone, which fulfills the requirements of cadaver trials and is very meaningful toward clinic applications. 2) A lot of experiments, including gesso bone and human bone experiments, are conducted to test the efficiency and accuracy of WATO-II. In order to test the effect of the surgery, we design the distal femoral appearance testing component to qualitative measure the cutting surfaces. Besides, we propose the three-dimensional digitizer based measurement algorithm to quantitative measure the mechanical axis. The experimental results show that WATO-II has a higher degree of surgical accuracy, which lay the firm foundation for clinical applications and will be very helpful for the clinical surgery in the future.3) When WATO-II finishes the total knee replacement, we should quantitatively evaluate the surgical result. Symmetric balance of the soft tissue around the knee has an important influence for the knee surgery. The soft tissue is a typical three-dimensional deformable structure, and computer vision based deformable object motion analysis methods are effective approaches for soft tissue analysis due to their characteristics of non-contact and high-precision. In this area, we propose a deformable motion analysis method based on L_∞and second-order cone programming (SOCP) using a stereo rig, a monocular method based on L∞and linear programming (LP), and a monocular method based on L2.4) Besides the soft tissue balance, another important evaluation criterion is to judge whether the patient's motion mode is normal. The knee consists of femur and tibia connected by the joint. This structure is a typical articulated-rigid structure. In fact, each joint of the human body, as well as the skeleton of the human, are all three-dimensional articulated-rigid structures. In this area, we propose a novel evolutionary algorithm called Probability Evolutionary Algorithm (PEA) which is based on quantum computing and quantum inspired evolutionary algorithm, and propose a method based on PEA for 3D human tracking.