Fully Biobased, Robust, and High-Conductivity Hydrogel for High-Fidelity Electrophysiological Monitoring and Deep Learning-Assisted Stroke Rehabilitation

enables an integrated multifunctional bioelectronic system. As a skin-conformal bioelectrode, it provides high-fidelity acquisition of electrophysiological signals (ECG, EEG, and EMG), achieving a high signal-to-noise ratio (24.3 dB for ECG) compared to commercial Ag/AgCl electrodes. When integrated with deep learning algorithms, the platform enables autonomous assessment of Brunnstrom stages for stroke rehabilitation with an accuracy of 97.31%, while a wireless telemedicine system supports remote diagnosis and personalized healthcare management. In parallel, the hydrogel functions as a highly stable strain sensor for real-time motion monitoring and precise gesture recognition, enabling intuitive control of prosthetic devices. Additionally, the hydrogel acts as a triboelectric nanogenerator electrode, yielding an open-circuit voltage of 72.1 V to power its own functions, while a microcontroller system supports wireless telemedicine and remote rehabilitation monitoring. This work presents an eco-friendly strategy for fabricating high-performance, biobased flexible electronics suited for health monitoring, telemedicine, and soft robotics.