Stress classification using K-means clustering and heart rate variability from electrocardiogram

Abstract

In this study, we propose a method to classify individuals under stress and those without stress using k-means clustering. After extracting the R and S peak values from the ECG signal, the heart rate variability is extracted using a fast Fourier transform. Then, a criterion for classifying the ECG signal for the stress state is set, and the stress state is classified through k-means clustering. In addition, the stress level is indicated using the R − Speak value. This method is expected to be applied to the U-healthcare field to help manage the mental health of people suffering from stress.