Artificial Intelligence-Powered Electronic Skin for Stress Monitoring

Stress and mental disorders have become a leading contributor to the global burden of disease, especially given that there is a sharp increase in anxiety and depression in the aftermath of the COVID-19 pandemic. While current stress assessment relies on subjective questionnaires and clinical surveys, there is an unmet need for continuous and precise quantification of stress response evaluations. While progress has been made in the development of wearable sensors and continuous health monitoring, these devices can only monitor a limited set of physical signals and suffer from poor operational stability in biofluids. To meet the demand for reliable long-term continuous monitoring and objective stress assessment, we developed a consolidated artificial intelligence-reinforced electronic skin (CARES) platform that deconvolutes the biological mechanism behind stress responses. Integrated with miniaturized iontophoresis module, microfluidic channels and system-level engineering, the platform is capable of monitoring three key vital signs as well as six molecular biomarkers in human sweat. To enable continuous long-term daily monitoring, we developed a general material strategy for biochemical sensors by introducing analogous composite materials into the sensor interface that achieves an unprecedented long-term sweat biomarker analysis of over 100 hours with high stability. The high performance of CARES platform was evaluated through continuous multimodal physicochemical monitoring over 24-hour real-life activities, and can differentiate physiological and psychological stressors, as well as quantification of psychological stress responses with a high accuracy and confidence level. Such fully-integrated CARES platform can be adapted to monitor a broad spectrum of disease diagnosis and pave the way for numerous practical wearable applications such as intelligent healthcare and personalized medicine.