Shared Control for Upper Extremity Rehabilitation in Virtual Environments

Abstract

Interest in the rehabilitation applications for robots has been increasing. For example, various devices have been developed to aid in reaching movements of stroke, traumatic brain injury (TBI), and spinal cord injury (SCI) patients. Typically these devices provided guided reaching movements for elbow and shoulder. The robotic aspect allows for repeatability, along with additional data for post-session analysis. To date, robotic rehabilitation systems with haptic feedback have not fully exploited the capabilities of a haptic display device. The simulators primarily focus on obeying the physical laws that govern such systems in order to re-create realistic environments for rehabilitative tasks, or the robotic devices are employed only for their ability to carry the impaired limb through various trajectories. This paper will present a novel active assistance paradigm for interactions in virtual environments displayed via haptic interfaces. The author’s recent research efforts have focused on the design of perceptual overlays in virtual environments that are active rather than passive. Passive virtual fixtures have been the primary perceptual overlay in haptics, and have been used extensively as “virtual rulers” in teleoperation environments to improve operator performance of pick-and-place tasks. Active assistance in the form of shared control between the haptic device and the human operator has the potential to elicit even better performance in virtual and remote environment interactions, and also has implications for improving training effectiveness. The intended applications include stroke rehabilitation and training for pilots, manufacturing, and surgery. A description of perceptual overlays and details of the shared control paradigm are presented, along with results from some preliminary experiments on shared control haptic assistance for training in virtual environments.