Breast cancer is the most frequently diagnosed cancer among women worldwide, and early detection is essential for improving survival outcomes. Magnetic Resonance Imaging (MRI) enables highly sensitive detection of breast lesions that require MRI-guided biopsy for definitive diagnosis. Limitations of conventional MRI-guided biopsy techniques have motivated the development of MRI-compatible robotic systems to improve needle placement.
This project focuses on the software integration and control of the MAMRI robotic system for MRI-guided breast biopsy. A ROS-based framework is developed to modularize robot control, communication, and motion planning, enabling safe and predictable robot positioning within the constraints of the MRI environment. Inverse kinematics and path planning are implemented to generate collision-aware positioning motions that can be previewed prior to execution. In addition, 3D Slicer is integrated as a ROS-connected graphical user interface, allowing user input for planning and supervising robot positioning while providing intuitive visualization of MRI data, robot state, and planned trajectories.
The system will be evaluated on breast phantoms inside an MRI scanner to assess targeting accuracy, repeatability, workflow time, and overall usability. By validating the framework under realistic clinical conditions, this research aims to narrow the gap between a research prototype and a system suitable for clinical use.
