A System for Robotic Heart Surgery that Learns to Tie Knots Using Recurrent Neural Networks

TLDR

Tying suture knots is a time‑consuming task in minimally invasive surgery, and existing solutions rely on replaying manually programmed trajectories, but knot tying requires a controller with internal memory that cannot be achieved with traditional feedforward neural nets or support vector machines. Automating knot tying could greatly reduce total surgery time for patients. We use recurrent neural networks with adaptive internal states, trained by the Evolino algorithm, to learn from surgeon‑provided trajectories. Long short‑term memory RNNs trained by Evolino significantly increase the efficiency of suture knot tying in MIS compared to preprogrammed control.

Abstract

Tying suture knots is a time-consuming task performed frequently during minimally invasive surgery (MIS). Automating this task could greatly reduce total surgery time for patients. Current solutions to this problem replay manually programmed trajectories, but a more general and robust approach is to use supervised machine learning to smooth surgeon-given training trajectories and generalize from them. Since knot tying generally requires a controller with internal memory to distinguish between identical inputs that require different actions at different points along a trajectory, it would be impossible to teach the system using traditional feedforward neural nets or support vector machines. Instead we exploit more powerful, recurrent neural networks (RNNs) with adaptive internal states. Results obtained using long short-term memory RNNs trained by the recent Evolino algorithm show that this approach can significantly increase the efficiency of suture knot tying in MIS over preprogrammed control.