Bedside robotic endoscopes render surgeons autonomous from assistants, potentially improving surgical outcome and decreasing costs. Why then have they not been widely adopted? We take a step back and first characterize classic (non-robotic) endoscope use through observations, literature and a domain expert interview. We review the literature on bedside robotic endoscopes and find that existing controls, individually, do not have the power to support both intended and appropriated endoscope uses. We thus explore combining controls to support this diversity of uses. Through an iterative cycle, we design and implement a multimodal and mixed-initiative technique that combines two user controls and one system control. Our evaluations confirm that individual controls do not satisfy the diversity of endoscope uses, and also that our technique indeed does so. Our work highlights the relevance of HCI research in the medical domain through robotic systems.

A laparoscope is a key element in Minimally Invasive Surgery (MIS) as it provides visual feedback to the surgeon. To overcome the drawbacks induced by its manual operation by a human assistant, it can be fixed on the end-effector of a robotic assistant teleoperated by the surgeon: laparoscope displacements are commanded through a master input interface (e.g., joysticks, voice control, etc.) and replicated accordingly at the slave level. In this approach, precision is of high importance to ensure a good operability by the surgeon. This is why the state-of-the-art relies on rigid laparoscope holders with high-gain PID position control, ensuring high static and dynamic precision. However, in the event of undetected obstacles, such stiff systems generate high forces that may cause harm to the patient. Rather, a compliant behaviour is desirable but it leads to a lack of precision when disturbances occur, such as the unknown friction between the trocar and the laparoscope. In this paper we present a “compliant-and-precise” laparo-scope holder with 4 active Degrees of Freedom (DoFs). Its design is based on cable transmission used for haptic interfaces, thus it exhibits very high backdrivability. The paper shows how an intelligent PID position controller can be used to compensate for unknown friction at the trocar while keeping very low PID gains and a satisfactory tracking precision.

Accurate endoscope localization is a key element for Computer-Assisted Laparoscopic Surgery (CALS). External localizers track the endoscope pose and reconstruct the camera geometry through a calibration matrix. Perspective-n-Points (PnP) methods estimate this matrix from a set of paired point coordinates, given by the simultaneous position of a spherical marker on the image (2D coordinates) and in the external localizer frame (3D coordinates). In this paper, the PnP calibration is restated as an iterative Perspective-n-Spheres (PnS) problem: the spherical geometry of the marker is taken into account when computing the projection of its 3D center into the 2D image plane. Using calibration residuals as evaluation matrix, the proposed PnS provides better results compared to the State-of-the-Art (SoA) PnP methods. Furthermore, results show that the calibration yields different estimates due to different camera poses in the working volume. This is assumed to result from a lack of absolute precision in the marker localization across the external localizer workspace.