Implementation of Trajectory-Based Attractive Virtual Fixtures for Surgical Knot Tying on a Simulated Surgical Robot

Abstract

Robot-Assisted Surgery (RAS) constitutes the cornerstone of minimally invasive procedures, enhancing precision and dexterity. The present thesis focuses on the implementation of a ROS Based attractive Virtual Fixtures (VFs) in a simulated version of the Da Vinci Research Kit (dVRK). For this implementation, V-REP (now named CoppeliaSim) has been used to support surgical knot tying tasks. The main objective is to develop a force-based guidance system that enhances instrument stability by dynamically applying virtual attractive forces to predefined surgical trajectories. Our approach integrates surgeme recognition, to identify the closest predefined surgical paths and guide dVRK’s robotic manipulators within optimal trajectories. Patient Side Manipulators (PSMs) are teleoperated using computer mouse and keyboard, allowing users to control robotic instruments. The integration of attractive forces results in dynamic deviation corrections when exceeding a proximity threshold. Furthermore, pulling the PSMs back into the intended knot_tying path while maintaining surgical precision. This method aims to assist operators in overcoming challenges related to hand-eye coordination, ensuring smoother execution of surgical tasks. Our system calculates the closest surgeme, from the JIGSAWS dataset to the last tool position if it deviates over a predefined threshold, to guide dVRK's PSMs back on trajectory. PSMs are teleoperated using computer mouse and keyboard. The integration of attractive virtual forces results in corrective adjustments when deviation exceeds a proximity threshold, pulling the PSMs back toward the knot-tying trajectory. This thesis is an introduction to VFs applied on a simulated surgical robot and tries to contribute to the field of haptic assistance in robotic surgery.