Medical robots for MRI guided diagnosis and therapy

Abstract

Magnetic Resonance Imaging (MRI) provides the capability of imaging tissue with fine resolution and superior soft tissue contrast, when compared with conventional ultrasound and CT imaging, which makes it an important tool for clinicians to perform more accurate diagnosis and image guided therapy. Medical robotic devices combining the high resolution anatomical images with real-time navigation, are ideal for precise and repeatable interventions. Despite these advantages, the MR environment imposes constraints on mechatronic devices operating within it. This thesis presents a study on the design and development of robotic systems for particular MR interventions, in which the issue of testing the MR compatibility of mechatronic components, actuation control, kinematics and workspace analysis, and mechanical and electrical design of the robot have been investigated. Two types of robotic systems have therefore been developed and evaluated along the above aspects. (i) A device for MR guided transrectal prostate biopsy: The system was designed from components which are proven to be MR compatible, actuated by pneumatic motors and ultrasonic motors, and tracked by optical position sensors and ducial markers. Clinical trials have been performed with the device on three patients, and the results reported have demonstrated its capability to perform needle positioning under MR guidance, with a procedure time of around 40mins and with no compromised image quality, which achieved our system speci cations. (ii) Limb positioning devices to facilitate the magic angle effect for diagnosis of tendinous injuries: Two systems were designed particularly for lower and upper limb positioning, which are actuated and tracked by the similar methods as the first device. A group of volunteers were recruited to conduct tests to verify the functionality of the systems. The results demonstrate the clear enhancement of the image quality with an increase in signal intensity up to 24 times in the tendon tissue caused by the magic angle effect, showing the feasibility of the proposed devices to be applied in clinical diagnosis.