An EMG signal processing system for control of an ankle-foot orthosis

This paper presents an electromyographic (EMG)based digital processing system for estimating foot direction of motion and force exerted from muscles of the human lower leg. This information can be used in the control frameworks of ankle-foot orthoses (AFOs). In essence, it is usually necessary to estimate the ankle's muscles voluntary contraction (VC) and user's intention of motion in order to control an AFO in a reliable and effective way. Consequently, a data processing system is developed that can designate these parameters from EMG signals. The voluntary contraction is calculated based on the root mean square (RMS) of EMG signals, and the direction of motion is estimated by using a feature-based classifier. The experiments are performed on five healthy, sitting subjects and the methodology consists of selecting discriminative features and using basic classifiers. The results of this work show that the direction of motion can be estimated in real-time with high accuracy.