The goal of this project is to design and develop a teleoperated master-slave surgical system that can potentially assist the physician in performing breast biopsy with a magnetic resonance imaging (MRI) compatible robotic system. MRI provides superior soft-tissue contrast compared to other imaging modalities such as computed tomography or ultrasound and is used for both diagnostic and therapeutic procedures. The strong magnetic field and the limited space inside the MRI bore, however, restrict direct means of breast biopsy while performing real-time imaging. Therefore, current breast biopsy procedures employ a blind targeting approach based on magnetic resonance (MR) images obtained a priori. Due to possible patient involuntary motion or inaccurate insertion through the registration grid, such approach could lead to tool tip positioning errors thereby affecting diagnostic accuracy and leading to a long and painful process, if repeated procedures are required. Hence, it is desired to develop the aforementioned teleoperation system to take advantages of real-time MR imaging and avoid multiple biopsy needle insertions, improving the procedure accuracy as well as reducing the sampling errors.

The design, implementation, and evaluation of the teleoperation system is presented in this dissertation. A MRI-compatible slave robot is implemented, which consists of a 1 degree of freedom (DOF) needle driver, a 3-DOF parallel mechanism, and a 2-DOF X-Y stage. This slave robot is actuated with pneumatic cylinders through long transmission lines except the 1-DOF needle driver is actuated with a piezo motor. Pneumatic actuation through long transmission lines is then investigated using proportional pressure valves and controllers based on sliding mode control are presented. A dedicated master robot is also developed, and the kinematic map between the master and the slave robot is established. The two robots are integrated into a teleoperation system and a graphical user interface is developed to provide visual feedback to the physician. MRI experiment shows that the slave robot is MRI-compatible, and the ex vivo test shows over 85%success rate in targeting with the MRI-compatible robotic system. The success in performing in vivo animal experiments further confirm the potential of further developing the proposed robotic system for clinical applications.