Interventional treatment is a novel therapy for cardiac and intravascular diseases. Compared with traditional open surgery, intravascular interventions have significant advantages such as small wound, less pain and short recovery time. The application of robotic assistance technology in interventional treatment brings substantial benefits, namely high precision, radiation exposure reduction and improved security. By taking the development of minimally invasive vascular interventional robot as research background, the study of this dissertation focuses on some key issues of interventional robot, such as robot system design, vessel segmentation, autonomous interventional surgery. The main contributions of this thesis are as follows: Firstly, the distributed structure of the robotic system is proposed, which employs the CANopen protocol to implement manipulator and delivery subsystem's synchronous motion control. The delivery device's mechanism is designed with considerations of clinical sterilization and guidewire installation. A force sensor, which is mounted on the delivery device to measure force applied to the guidewire, improves the surgical security. Secondly, a dual-loop manipulator compliant controller is designed. The outer loop is an impedance controller which translates the force measured from 6 axis force sensor into desired speed in joint coordinates; The inner speed tracking controller is designed with the ability to suppress the errors caused by external perturbation and dynamic model errors. Manipulator employing this compliant strategy is able to detect and follow the surgeons' desired direction, which improve surgeons' agility. Thirdly, we proposed the delivery subsystem's control strategy and optimized the control flow, and then analyzed various errors' transmission in delivery subsystem. According to the analytical results, a function based on vessel width is proposed to improve delivery precision in small vessels. Fourthly, we presented a template matching algorithm for vessel segmentation. This algorithm, which consists of template predicting, template matching and a branch detection algorithm, is able to detect small vessels and vessels in low contrast background, offering surgeons more precise visualization. Fifthly, the autonomous robotic intervention is studied and an image-based algorithm is proposed. This algorithm employs preoperative CT data and intraoperative electromagnetic tracking system to control robotic syst...
修改评论