An EMG-driven neuromuscular interface for human elbow joint

This paper presents the development of a neuromuscular interface for an exoskeleton to assist the elbow joint. The interface uses electromyographic (EMG) signals obtained from the biceps and triceps to predict elbow flexion and extension movements. These movements occur in the sagittal plane and the effects of forearm weight have been incorporated. The interface uses a physiological-model-based approach to convert the EMG signals to a joint displacement and this is based on Hill-type muscle models that have traditionally been used in clinical applications or for diagnosing and managing neurological and orthopaedic conditions. Simulation results have been obtained for the performance of the interface on pre-recorded data. While the general trend of movement was correctly identified by the interface, further experiments are required to quantify the accuracy and determine real-time performance capabilities.