Neurosurgical robot design and interactive motion planning for resection task

Abstract

This paper presents the modeling of a new mini-invasive neurosurgical resection robot. This robot aims to help to remove brain tumors and is incorporated into a multi-robot neurosurgical system. We focus especially on the resection task. The robot is composed of three serial bending modules actuated by wires (cables) and uses an additional translation. The redundancy of the robot (seven degrees of freedom) allows more dexterity for the resection task. A kinematic (geometric) model is built distinguishing the structure and the actuation models, and integrating the influence of the wires into the kinematic behaviour of the robot. A method for kinematic redundancy handling is defined and assigns different parts of the robot to different tasks. An interactive path planning based on arbitrary sequence of elementary procedures assembled by the neurosurgeon is proposed and the yielding procedures are computed from the kinematic model.