| Abstract Scope |
Respiration and heartbeat are critical physiological indicators in medical diagnostics, yet existing methods are often cumbersome or rely on manual assessment, limiting their effectiveness. The COVID-19 pandemic has underscored the urgency of robust respiratory monitoring solutions, especially with the widespread adoption of masks. This work introduces a machine learning-enhanced, mask-type wireless wearable system designed for simultaneous monitoring of heartbeat and respiration. The device integrates thermoelectric bulks embedded into the mask, cloth electrodes, and an optical sensor to track these vital signs in real-time. A key innovation is the incorporation of deep learning algorithms, specifically an XGBoost classifier, to enable multivariate fusion recognition of distress phrases and abnormal heartbeat patterns. This integration of machine learning significantly enhances the system's ability to detect and respond to critical health events, providing a groundbreaking tool for continuous and effective patient monitoring in medical and daily-life scenarios. |