A Haptic Object to Quantify the Effect of Feedback Modality on Prosthetic Grasping

Abstract

Upper limb prostheses do not provide proprioceptive cues to their users, which impedes performance in grasp and manipulation tasks. Various sensory substitution methods have been developed to provide prosthesis users with a sense of the device's position, orientation, and terminal device aperture. However, the impact of these technologies on manipulation dexterity has not been systematically assessed and these technologies have not been widely adopted by prosthesis users. Researchers have previously used visually rendered virtual objects and virtual prostheses, which do not provide cues associated with physical interactions involved in grasping and manipulation. Other studies use a physical object and a physical prosthesis, but these methods lack the benefit of programmable conditions, systematic experiment setup, and direct evaluation of performance in addition to issues of consistent application across different types of prostheses. In order to address these concerns, we suggest a new method to evaluate different types of feedback with the use of a haptic device that renders physical properties through motors to a hand or a physical prosthesis. Ten users without amputation were asked to match hand aperture to a target using various feedback modalities (visual, vibration, and soft force-wall feedback). Participants interacted with the haptic device either through their natural hand or using a body-powered prosthetic emulator. In addition, three prosthesis users were recruited to complete the same task, each using their own prosthesis. Trials performed under conditions with visual and vibration feedback showed similar improvements in position accuracy, compared to trials without feedback.