摘要
In view of the need for real-time, multimodal respiratory health monitoring, we present a wearable sensor platform (BreathFIT) that simultaneously measures lactate and electromyography (EMG) signals for assessment of diaphragm function. By virtue of integrating liquid metal-polydimethylsiloxane (LM-PDMS) flexible sticky conductors for EMG acquisition with a lactate screen-printed electrode (LSPE), the platform enables concurrent measurement of muscle electrical activity and metabolic byproducts (lactate) from the diaphragm. Data are captured and visualized in real time via custom software, enabling machine learning models (LSTM, DNN, SVM, Random Forest) to classify diaphragm activity patterns during rest, exercise, and sleep. Owing to its integrated design, the system exhibits stable electrochemical performance, exceptional mechanical compliance (300 % stretchability), excellent skin conformality (39 N cm −1 adhesion strength), and high accuracy in distinguishing diaphragm states. Consequently, all models demonstrate high classification accuracy, with Random Forest proving the most effective at capturing time-series features (accuracy = 92.19 %). The dual-modality platform uniquely integrates electrical and biochemical sensing with intelligent analysis. This configuration enables high spatiotemporal-resolution monitoring of respiratory muscle activity. Consequently, the system opens a new avenue for wearable multimodal health monitoring, providing significant potential for both clinical and at-home respiratory health management.
危险品化学品经营许可证(带存储)