Towards MRI-Guided and Actuated Tetherless Milli-Robots

Image-guided and robot-assisted surgical procedures are rapidly evolving due to their potential to improve patient management and cost effectiveness. Magnetic Resonance Imaging (MRI) is used for pre-operative planning and is also investigated for realtime intra-operative guidance. A new type of technology is emerging that uses the magnetic-field gradients of the MR scanner to maneuver ferromagnetic agents for local delivery of therapeutics. With this approach, MRI is both a sensor and forms a closed-loop controlled entity that behaves as a robot (we refer to them as MRbots). The objective of this thesis is to introduce a computational framework for preoperative planning using MRI and modeling of MRbot maneuvering inside tortuous blood vessels. This platform generates a virtual corridor that represents a safety zone inside the vessel that is then used to access the safety of the MRbot maneuvering. In addition, to improve safety we introduce a control that sets speed based on the local curvature of the vessel. The functionality of the framework was then tested on a realistic operational scenario of accessing a neurological lesion, a meningioma. This virtual case study demonstrated the functionality and potential of MRbots as well as revealed two primary challenges: real-time MRI (during propulsion) and the need of very-strong gradients for small MRbots for maneuvering inside narrow cerebral vessels. Our ongoing research focuses on further developing the computational core, MR tracking methods, and on-line interfacing to the MR scanner.