Gaze Contingent Cartesian Control of a Robotic Arm for Laparoscopic Surgery.

Abstract

This paper introduces a gaze contingent controlled robotic arm for laparoscopic surgery, based on gaze gestures. The method offers a natural and seamless communication channel between the surgeon and the robotic laparoscope. It offers several advantages in terms of reducing on-screen clutter and efficiently conveying visual intention. The proposed hands-free system enables the surgeon to be part of the robot control feedback loop, allowing user-friendly camera panning and zooming. The proposed platform avoids the limitations of using dwell-time camera control in previous gaze contingent camera control methods. The system represents a true hands-free setup without the need of obtrusive sensors mounted on the surgeon or the use of a foot pedal. Hidden Markov Models (HMMs) were used for real-time gaze gesture recognition. This method was evaluated with a cohort of 11 subjects by using the proposed system to complete a modified upper gastrointestinal staging laparoscopy and biopsy task on a phantom box trainer, with results demonstrating the potential clinical value of the proposed system.