Effectiveness of supplemental grasp-force feedback in the presence of vision

Previous studies have shown that supplemental grasp-force feedback can improve control for users of a hand prosthesis or neuroprosthesis under conditions where vision provides little force information. Visual cues of force are widely available in everyday use, however, and may obviate the utility of supplemental force information. The purpose of the present study was to use a video-based hand neuroprosthesis simulator to determine whether grasp-force feedback can improve control in the presence of realistic visual information. Seven able-bodied subjects used the simulator to complete a simple grasp-and-hold task while controlling and viewing pre-recorded, digitised video clips of a neuroprosthesis user's hand squeezing a compliant object. The task was performed with and without supplemental force feedback presented via electrocutaneous stimulation. Subjects had to achieve and maintain the (simulated) grasp force within a target window of variable size (+/- 10-40% of full scale). Force feedback improved the success rate significantly for all target window sizes (8-16%, on average), and improved the success rate at all window sizes for six of the seven subjects. Overall, the improvement was equivalent functionally to a 35% increase in the window size. Feedback also allowed subjects to identify the direction of grasp errors more accurately, on average by 10-15%. In some cases, feedback improved the failure identification rate even if success rates were unchanged. It is thus concluded that supplemental grasp-force feedback can improve grasp control even with access to rich visual information from the hand and object.