Selection of Neural Oscillatory Features for Human Stress Classification with Single Channel EEG Headset

Abstract

A study on classification of psychological stress in humans using electroencephalography (EEG) is presented. The stress is classified using a correlation-based feature subset selection method that efficiently reduces the feature vector length. In this study, twenty-eight participants are involved by filling in the perceived stress scale-10 (PSS-10) questionnaire and their EEG is also recorded in closed eye condition to measure the baseline stress. The recorded data is labelled on the basis of the stress level that is indicated by the participant's PSS score. The feature selection method has shown that, among the EEG oscillations, low beta, high beta, and low gamma are the most significant neural oscillations for classifying human stress. The proposed method not only reduces the time to build a classification model but also improves the classification accuracy up to 78.57% using a single channel wearable EEG device.

A Review on Mental Stress Assessment Methods Using EEG Signals

“…There was an obvious variation in the treated frequency ranges. Thus, the highest accuracy was acquired when dealing with seven bands where they got 85.20% for SVM [ 108 ] comparing with the three bands 78.57% [ 54 ] and four bands 83.33% [ 106 ] that have used same classifier. Besides, the lowest performance was related to two frequency fields with 71.4% accuracy with NB classifier.…”

“…Pearson’s correlation-based captures linear, time-domain dependencies among EEG signals. It could be found over a single epoch or over several epochs, and it is calculated using the Pearson’s correlation coefficient, cross-covariance, and auto-covariance of EEG signals [ 54 ]. Therefore, increasing the value for the Pearson correlation coefficient from (−1) to (1) indicates intense connections between brain regions.…”

“…Therefore, increasing the value for the Pearson correlation coefficient from (−1) to (1) indicates intense connections between brain regions. In particular, this technique has been used by study [ 54 ] to reduce feature vectors and computational time, and to improve accuracy of SVM classifiers in detecting human stress. The main interest of such features is the high performance while reducing dimensionality of the EEG data set [ 55 ].…”

“…In particular, PSD is useful in describing stochastic process of the signal and evaluating short data records [ 78 ]. There are several methods applied to estimating the PSD, for example, fast Fourier transform (FFT), Welch, Burg, Yule walker, welch method and periodogram [ 54 ]. Several studies have demonstrated the effectiveness of using PSD to estimate the level of stress.…”

Electroencephalography Correlates of Well-Being Using a Low-Cost Wearable System

“…These applications include brain monitoring in naturalistic settings and in real-time (Hu et al, 2015;Jebelli et al, 2017), brain-computer interfaces (BCI; Park et al, 2020), neurofeedback interventions (Angelakis et al, 2007;Quaedflieg et al, 2016;Brandmeyer and Delorme, 2020a), neuromarketing (Cartocci et al, 2018;Ramsøy et al, 2018), or neuroaesthetics research (i.e., the science studying the biological underpinnings of aesthetic experience; Cheung et al, 2019;Cartocci et al, 2021). While these EEG systems can have inferior hardware capacities than conventional ones, recent technological and algorithmic advancements make the detection and measurement of mental states increasingly reliable (Wu et al, 2017), with as few as a single EEG channel (Umar Saeed et al, 2018;Arpaia et al, 2020;Mahmoodi et al, 2021). Additionally, these systems can easily combine other physiological measures such as electrocardiography (ECG) or galvanic skin response (GSR) to improve the efficacy of mental states detection (e.g., Ahn et al, 2019).…”