Collision detection and untangling for surgical robotic manipulators

Abstract

BACKGROUND: Robotic-assisted minimally invasive surgery provides several advantages over traditional surgery; however, it also has several drawbacks, such as possible collisions between the robotic arms and a limited field of view. METHODS: A generic method for tracking the configuration of a surgical manipulator in real time is presented. It is coupled with a collision detection and dynamic simulation algorithm, allowing the operator to detect collisions between robotic arms before they happen and presenting the best possible untangling direction to get out of collisions. RESULTS: Our algorithm successfully tracks the configuration of the Zeus surgical system and accurately detects possible collisions in real time. A pilot study on our system proved its efficiency in reducing the duration and severity of collisions, at the price of longer completion times. CONCLUSION: Our system helps alleviate the collision problem by reducing the time in collision.