A Simulated Intelligent Pixelated Electrode Array for Surface Electromyography Sensors

Surface electromyography (sEMG) is a noninvasive diagnostic tool for identifying muscle diseases, which is also utilized in portable smart devices to recognize body mass compositions. Human skin tissue composition varies from person to person. Therefore, the strength of muscle stimulation that reaches the skin surface varies with skin formation, resulting in a variation of signal-to-noise ratio (SNR). Electrodes used in sEMG have inherent filtering capabilities and the cutoff frequencies depend on the dimension of the electrodes. In this article, we propose a novel electrode array composed of tiny electrodes, mimicking the arrangement in high-density sEMG, that can modulate electrode size and leverage the natural filtering property of electrodes to maximize the SNR. The arrangement is similar to the arrangement of image pixels on a digital screen and hence called “pixelated.” The proposed scheme is very adaptive to varied skin composition as well as body location and hence behaves like an intelligent agent. We used a statistical SNR prediction algorithm, which validated the proposed idea. While determining the ideal electrode layout, the suggested concept significantly reduces the signal sample duration and can avoid additional computational overhead once the optimal arrangement is identified. Moreover, it offers a substantial advantage over traditional signal processing techniques, which need constant processing to filter out noise and improve the intelligent signal with digital filters and transceiver amplifiers. Because of truncated processing cost, the proposed array is suitable for wearable devices driven with low-power embedded microcontrollers, where high computational requirement often proves to be prohibitive.