Timely mental stress detection can help to prevent stress-related health problems. The aim of this study was to identify those physiological signals and features suitable for detecting mental stress in office-like situations. Electrocardiogram (ECG), respiration, skin conductance and surface electromyogram (sEMG) of the upper trapezius muscle were measured with a wearable system during three distinctive stress tests. The protocol contained stress tests that were designed to represent office-like situations. Generalized Estimating Equations were used to classify the data into rest and stress conditions. We reached an average classification rate of 74.5%. This approach may be used for continuous stress measurement in daily office life to detect mental stress at an early stage.
展开▼