Next-Generation 6-DoF Handheld Manipulator for Microsurgery

Abstract

Robotic assistance is used today in a variety of surgeries as a means of precise, dexterous, and minimally-invasive manipulation. However, practical use in microsurgical environments such as vitreoretinal surgery remains a challenge for the most common mechanically-grounded robotic platforms. Microsurgery requires micron-level accuracy and the ability to manipulate with interaction forces in millinewtons. Vitreoretinal surgery specifically requires a line of sight through the lens of the eye and maintaining a remote center of motion (RCM) at the incision point in the eye.

A handheld manipulator, Micron, has been presented to address the unique challenges of microsurgery. It enables the manipulator to scale down a surgeon's motion, actively cancel their involuntary tremor, and resist moving the tooltip into unsafe areas. Since the manipulator is handheld, it does not significantly obstruct the operating room and allows the surgeon to still feel haptic feedback during surgery. The manipulator is also capable of moving around an RCM at the sclerotomy. However, the manipulator's mechanical design necessitated unreliable motors that have a low stall force and a more reliable iteration was needed.

This thesis describes the construction of a new iteration of Micron based on the KIST microsurgical robot and integrated with our group's control systems and optical sensor. It begins by investigating the applicability to Micron of a design based on parallel continuum manipulators. The kinematics and dynamics of a parallel continuum manipulator are simulated and a further simulation is built to evaluate the stiffness of the designs. Although the design ultimately proves infeasible, the analysis settles the question for future research. Two simulations of the rigid-link Micron, built in Gazebo and PyBullet, are also created and evaluated as first steps toward a full kinematics and dynamics simulation of the manipulator. The PyBullet simulation is simplified and automated to pick up a small object on a model of the retina. Most importantly, the physical manipulator is constructed, including all its mechanical and electrical parts, as well as firmware to integrate it into our group's control software and optical sensor. The manipulator is actuated and its pose measurements are analyzed.