High-Fidelity Recording of EMG Signals by Multichannel On-Skin Electrode Arrays from Target Muscles for Effective Human–Machine Interfaces

Human hands are the most dexterous parts of the human body where the finger movements are mainly controlled by several specific forearm muscles. The accurate acquisition of surface electromyography (sEMG) signals from these target muscles is essential for hand gesture recognition widely applied in human–machine interface (HMI) systems. However, most of the existing sEMG sensors are designed as single bipolar electrode pairs or orthogonal electrode arrays, ignoring the irregular spatial distribution of slender forearm muscles, which limits their performances in signal acquisition. Herein, we propose customized four-channel electrode arrays where the electrode pairs are placed in accordance with the position and orientation of target muscles. By selecting materials with excellent properties for on-skin devices, the fabricated electrodes achieve low skin–electrode impedance and record sEMG signals with a high signal-to-noise ratio (SNR). Owing to the customized design, our electrode arrays can cover more muscles and record higher-quality multichannel sEMG signals than orthogonal arrays under the same condition, enhancing the accuracy of hand gesture classification. The customized electrode arrays proposed in this study are promising for various HMI applications in which EMG signals or hand gestures are adopted as control signals.