A Wearable Foot-Based IMU-sEMG Controller for Real-Time Upper-Limb Prosthetic Control

Introduction: Advanced upper-limb prostheses offer multiple degrees of freedom, but residual-limb-dependent control methods limit their use. This study presents, to our knowledge, the first upper-limb prosthetic controller to co-locate a foot-mounted inertial measurement unit (IMU) and a toe-region surface electromyography (sEMG) sensor in one wearable device, supporting continuous motion and discrete commands while stationary and walking. Methods: The developed IMU architecture was validated using an IMU-only controller combining a foot-mounted IMU and a two-button shoe insole. Eleven participants completed a real-world baseline, 10 virtual-reality (VR) training sessions, and a real-world transfer session using below-the-elbow control mapping. Then a hybrid controller with the same foot-orientation mappings, command, and communication structures was developed, replacing the insole buttons with one toe-region sEMG sensor and upgraded walking-state detection. IMU-only results supported the learnability and functionality of the hybrid system’s IMU component. The hybrid controller was evaluated with three users through object manipulation, indoor walking, and real-time control of a MATLAB/Blender shoulder-level animated arm. Results: IMU-only participants improved their performance through training in VR, increasing block-sorting speed without sacrificing accuracy while maintaining or improving real-world performance. All hybrid-controller users successfully generated intentional IMU and sEMG commands and completed their tasks with the prosthesis. While walking, the controller suppressed gait-related IMU signals, preserved deliberate sEMG commands through burst-based classification, and re-zeroed its orientation afterward. Unintended walking triggers were absent during the short trials, and missed intentional triggers were rare. The shared architecture enabled real-time control of the physical below-the-elbow prosthesis and a higher-degree-of-freedom, shoulder-level animated arm. Conclusions: These preliminary findings demonstrate a novel foot-based IMU-sEMG architecture for prosthetic control while stationary and walking, overcoming previous control limitations. Larger controlled studies and physical shoulder-level testing will be followed to evaluate reliability, usability, and long-term performance.