GPC-based Teleoperation For Delay Compensation And Disturbance Rejection In Image-guided Beating-heart Surgery

Beating-heart surgery is not currently possible for most surgical procedures because it requires superhuman skill to manually track the heart’s motion while performing a surgical task. However, if we could make the surgical tool track the motion of the point of interest (POI) on the heart, then, with respect to the surgical tool tip the POI would appear stationary. Such a system can be created with the assistance of a surgical robot. The teleoperated surgical tool can be controlled to track the combination of the heart’s motion and the surgeon’s motion as input through a separate user console. When developing such a system, the motion of the heart, which is required for position control of the surgical robot, is found in ultrasound images where the image acquisition introduces delays of approximately 40 ms and image processing further increases this delay. Directly using this delayed position measurement in the feedback control loop can lead to instability and poor tracking. The generalized predictive controller used in this work compensates for this time delay despite the large disturbances with velocities up to 210 mm/s and accelerations up to 3800 mm/s^2 introduced by the moving heart. In this study, the surgical tool is able to track the heart’s motion with a mean error of 8.1 mm and a normalized integrated squared error 91 mm^2 when the amplitude of the simulated heart motion is 20 mm.