Biomimetic, Antisweat Flexible Hydrogel Sensor With Deep Learning for ECG Monitoring, Fatigue Assessment, and Prevention

Chronic fatigue triggers multi-system symptoms and cognitive impairments, increasing the risk of cardiovascular and metabolic diseases. Although the electrocardiogram (ECG) offers real-time, noninvasive detection of physiological abnormalities associated with chronic fatigue, traditional electrodes are susceptible to interference from sweat and motion artifacts, and may cause skin irritation, reducing the accuracy of assessments. Here, we introduce a flexible and long-term sweat-resistant ECG acquisition system. By integrating bio-inspired polydopamine (PDA) hydrogel electrodes with flexible microcircuitry, the system enables unobtrusive, continuous acquisition of heart rate variability (HRV) features for robust fatigue assessment. The developed hydrogel electrode achieves long-term stability (> 30 days), retaining 90% conductivity after 500 mechanical cycles, and demonstrates over 40% improvement in signal-to-noise ratio (SNR) under humid conditions. Its excellent mechanical properties, self-adhesion, and light transmittance make it suitable for human-machine interaction, including real-time monitoring of dynamic motion signals, characteristic pulse waveforms, and handwriting trajectory recognition. Furthermore, by integrating flexible microcircuits and robust deep learning algorithms, fatigue assessment can achieve a recognition accuracy of 88.7% and provide early warnings. Through the synergistic integration of bio-adhesion, environmental adaptability, and signal fidelity, the flexible biomimetic ECG acquisition sensor provides strong support for personal health risk assessment, early warning, and precise intervention, thereby promoting the intelligent and personalized development of health management.