Stress has almost become a daily experience, mainly because of its eclectic nature – physical, psychological, emotional, social, etc. There have been attempts to detect and quantify stress. However, only physiological measurements can be used for an accurate assessment. There has been an attempt for stress detection through continuous monitoring of Electro Dermal Activity (EDA).
Stress when unmonitored can prove to be deleterious to a one’s health and can lead to chronic diseases.
HTIC has developed a machine learning model to classify EDA into stress and non-stress regions and also detect regions affected by motion artefacts. A clinical validation study was conducted to establish the efficacy of the algorithm in detecting stress in a natural environment. The study was conducted on 30 participants that were subjected to Trier Social Stress Test (TSST), EDA and accelerometer data were recorded using a wrist worn device. Based on recorded stress protocol timeline, datasets containing stress and non-stress periods were segmented and manually tagged for model training.