Evaluation and Restriction of Interaction Forces with Softbodies for Minimally-Invasive Robotic Surgery

Abstract

This thesis aims to address the problem of difficult sensor integration for haptic feedback in minimally invasive robotic surgery by modeling soft tissue through the use of virtual fixtures. Instead of directly feeding forces recorded by a sensor back to the surgeon they are used to update a model of the tissue or surgical suture. This model is based on at run-time parameterizable virtual fixtures implemented by Hagmann et al. [1], which render forces on a haptic input device. The surgical environment is abstracted with the Lübecker Toolbox Nylon Twist training pad [2]. Virtual Fixtures are combined into units of two, with one modeling interaction forces and the other preventing critical levels of force. The units have multiple configurations for different surgical scenarios such as suture or the pulling of soft tissue, represented in the training pad. Different sources for force data are examined in terms of their usability for updating this works model. A pilot study indicates a reduction in workload for all configurations when integrating fixture units and a preference of users for tasks with haptic feedback.