Design of Unobtrusive Wearable Mental Stress Monitoring Device Using Physiological Sensor

Abstract

Stress is recognized to be the most prevalent life risk factor all over the world. Chronic stress can elicit various mental disorders and increase the risk of cardiovascular disease. However, stress measurement from psychiatrist is instantaneous, qualitative and quarrelsome. Furthermore, there is a need to measure the stress level ubiquitously to prevent the chronic stress from happening. This study proposed an inconspicuous stress monitoring system using physiological sensors with accurate algorithm to measure the stress level of the user. We fabricated a stress monitoring patch prototype with mobile app user interface and conducted experiment with 50 participants to classify the stress and relax data distribution from three physiological signals namely heart rate, skin temperature, and galvanic skin response. We applied support vector machine algorithm and K-means clustering to classify the obtained training data and index the stress level of the user which resulted in the overall accuracy of 91.26 also has feature to alert the users if their stress level is more than 80 assist the users in alleviating their stress.