Catalyst Awardee

Project Description

Machine Learning-Driven Multimodal Wearable Biosensor for Stress Monitoring

Wei Gao, PhD | California Institute of Technology
Competition Sponsor: US National Academy of Medicine
Awardee Year: 2022

High levels of stress and anxiety can significantly affect human performance and can substantially reduce life expectancy by several years. Early detection and classification of the severity of stress allow for timely intervention and can be used to foster mental wellness proactively. However, current approaches for stress assessment are largely limited to subjective questionnaire-based scales, and to date, there is no existing method that accurately and objectively monitors stress. We hypothesize that molecular biomarker levels monitored continuously by non-invasive sweat analysis along with the key vital sign monitoring, when coupled with a machine learning approach, will provide crucial insight into mental health status and can be used for accurate and dynamic stress and anxiety assessment toward timely intervention. We propose a holistic hardware/software solution based on a multimodal wearable sensing platform to achieve dynamic stress and anxiety assessment. Our strategy is to simultaneously monitor the molecular analytes in human sweat including stress hormones (i.e., cortisol, adrenaline, and norepinephrine), glucose, lactate, uric acid, Na+, K+, pH, sweat rate, and key vital signs (i.e., temperature, GSR, blood pressure pulse waveforms) using a wearable multimodal sensing platform which can autonomously induce and analyze sweat across activities. We plan to validate this wearable patch in human studies and apply machine-learning data analysis for dynamic stress classification and prediction. We anticipate that the algorithm generated from this study will have a major impact in personalized stress assessment, and potentially many other major health conditions.

To learn more about this proposal, email

Sign up for updates