A Wearable Multi-Sensor System for Flexible Wearing and Robust Monitoring of Muscle Activity

Muscle activity monitoring plays a vital role in diverse fields, including clinical examination, rehabilitation assessment, and human-robotics interaction. However, the quality of acquired physiological signals often suffers due to limitations in sensor flexibility, robustness, and resistance to interference, which hampers the realization of long-term muscle activity monitoring. In this study, we introduced a novel wearable, flexible monitoring system, comprising a Magnetic-Inertial Measurement Unit (MIMU) and surface electromyography (sEMG) sensor. The designed flexible electrodes exhibited superior performance with low contact impedance and high signal-to-noise ratio. Additionally, we proposed an adaptive filtering algorithm based on a finite-state machine framework. This algorithm effectively compensated for orientation information by utilizing an electromagnetic interference observer and dynamically filtered electromyographic information through automated notch filter adjustments, enhancing the system's robustness against electromagnetic disturbances. Furthermore, validation experiments were conducted on five subjects across different muscle groups. The results, obtained from 12 hours of continuous monitoring with measurements taken every two hours, showcased consistent low contact impedance (60.32±12.15kΩ) and high signal-to-noise ratio (17.64±0.53 dB) of the designed flexible sensors. Notably, the system successfully improved posture-related information and dynamically filtered sEMG signals in environments with random magnetic field interference. In summary, our proposed flexible wearable monitoring system demonstrates promising capabilities in acquiring high-quality sEMG-orientation data, offering reliable and superiorquality information for applications in human-robot interaction.