Toward Portable Hybrid Surface Electromyography/A-Mode Ultrasound Sensing for Human–Machine Interface

Abstract

It is evident that non-invasive muscle-based human–machine interface (HMI) has been the research focus of human–machine interaction. To improve the performance of muscle-based HMI, it is significantly important to obtain electrophysiological and morphological changes of muscle contraction. However, there is still lacking of solution to present electrophysiological and morphological information of the same muscle at the same time. Surface electromyography (sEMG) can reflect the electrical activity of functional muscle contraction and A-mode ultrasound (AUS) can monitor the morphological structure of active muscle, both in non-invasive manners. This paper proposes a portable hybrid sEMG/AUS system for HMI. The system consists of composite sensor armband and signal acquisition modules, where the former achieves arrangement of two kinds of sensors at the same muscle position and the latter enables the simultaneous acquisition of sEMG and AUS signals. The hardware evaluation experiment proves that the system can provide high-quality signals in respect to signal-to-noise ratio (SNR) and time–frequency characteristics. Furthermore, the hand gesture recognition experiment validates the complementarity between sEMG-based and AUS-based HMI, since the recognition accuracy of hybrid sEMG/AUS feature is significantly improved by 4.85% ( ${p}$ = 0.0095) and 20.6% ( ${p} ) compared to the results of ultrasound features and sEMG features, respectively.