Design, Modeling and Optimization of a Magnetic Resonance Conditional 3-RRR Spherical Parallel Robot for Neurosurgery

Abstract

In neurosurgery, magnetic resonance (MR) imaging is extensively utilized for preoperative diagnosis and postoperative evaluation due to its superior soft tissue contrast. However, the strong magnetic field poses a challenge to the real-time utilization of MR for intraoperative navigation. To facilitate neurosurgery in the MR environment, this paper develops a MR conditional robot featuring nonferrous materials and ultrasonic motor actuation. The robot consists of a 3-degree-of-freedom (3-DOF) translational module and a 3-DOF remote center of motion (RCM) module. The RCM module incorporates a 3-RRR spherical parallel mechanism. The mechanical design and kinematic modeling of the RCM module is completed. This paper further conducts the optimization for the RCM module. Additionally, a path-planning algorithm, focusing on the maximization of dexterity, is introduced, and the feasible workspace of the optimized RCM module is evaluated. A prototype is fabricated, and the orientation repeatability of the RCM module is measured to be 0.055±0.0016°, and the absolute orientation error is 2.05±0.019°. Needle insertion experiments are performed on an agarose phantom to evaluate the feasibility of the robot. The impact on signal-to-noise ratio in MRI images caused by the robot is less than 4%, indicating a highly promising applicability in MR conditional neurosurgery.