Autonomous Operation in Surgical Robotics

摘要

The article focuses on developing an algorithm for automation based on stereo computer vision and dynamic registration in a surgical robotic context. The performance of the algorithm was further tested experimentally utilizing the block transfer task which corresponds to tissue manipulation as designed by Fundamentals of Laparoscopic Surgery (FLS). The surgical robotics field as a whole progresses towards the reduction of invasiveness limiting the trauma at the periphery of the surgical site and increase of semi-autonomous operation while positioning the surgeon as a decision maker rather than as an operator. The autonomous FLS task is implemented successfully and tested experimentally with the Raven II surgical robot system. The data indicate that the autonomous operational mode has better overall performance and limited tool-environment interaction compared with the human teleoperation mode. Surgeon’s intention may also be extracted from a database that may lead to seamless switching between the human operator and the autonomous system and in that sense, it may allow the autonomous algorithm to cope with more complex surgical environments.