Osseoperceptive EMG Feedback Improves Grasp Force Control in Myoelectric Prosthetic Hands
Cristian Felipe Blanco-Diaz1,2, Strahinja Dosen3, Leonardo Cappello1,2
1The BioRobotics Institute, Scuola Superiore Sant'Anna, Pisa, Italy.
Abstract:
Restoring sensory feedback is essential for achieving accurate and natural control of myoelectric prosthetic hands, yet most clinical systems remain open loop. Osseoperception, the transmission of vibrotactile and auditory sensations through bone-conducted mechanical stimulation, has emerged as a promising approach for delivering artificial sensory feedback. In this study, osseoperceptive electromyography (EMG) feedback is used to convey the user's myoelectric signal through bone conduction, enabling predictive regulation of grasp force. A single bone transducer placed on the elbow provides three encoding schemes-vibrotactile, auditory, and combined vibrotactile-auditory-evaluated during a functional force-matching task inspired by the Box and Block test. Thirteen non-disabled participants and one transradial amputee performed grasp-and-transfer trials across five discrete force levels, with and without feedback. The main outcome measure was the success rate in producing the correct target force. In non-disabled participants, auditory-only encoding significantly improved force-control accuracy compared to no feedback (71% vs. 50%), whereas vibrotactile-only and combined encoding showed positive but non-significant trends (both 60%). The amputee participant achieved the best performance with vibrotactile encoding. These findings demonstrate the feasibility of osseoperceptive EMG feedback for predictive grasp-force regulation using a single lightweight actuator and support its potential as a non-invasive strategy for closed-loop prosthetic control.


