Stress Classification Using Brain Signals Based on LSTM Network

Abstract

The early diagnosis of stress symptoms is essential for preventing various mental disorder such as depression. Electroencephalography (EEG) signals are frequently employed in stress detection research and are both inexpensive and noninvasive modality. This paper proposes a stress classification system by utilizing an EEG signal. EEG signals from thirty-five volunteers were analysed which were acquired using four EEG sensors using a commercially available 4-electrode Muse EEG headband. Four movie clips were chosen as stress elicitation material. Two clips were selected to induce stress as it contains emotionally inductive scenes. The other two clips were chosen that do not induce stress as it has many comedy scenes. The recorded signals were then used to build the stress classification model. We compared the Multilayer Perceptron (MLP) and Long Short-Term Memory (LSTM) for classifying stress and nonstress group. The maximum classification accuracy of 93.17% was achieved using two-layer LSTM architecture.

A scalable and robust system for Audience EEG recordings

“…Other studies have also demonstrated that it can measure ERP components of later visual processing (Krigolson et al, 2021; Krigolson et al, 2017). Multiple studies have demonstrated that it can be used to measure various aspects of brain function related to Stress (Asif et al, 2019; Phutela et al, 2022), Attention (Vortmann et al, 2022), Mindfulness (Hawley et al, 2021; Hunkin et al, 2021), Stroke (Wilkinson et al, 2020), Drowsiness (LaRocco et al, 2020), Emotion Classification (Raheel et al, 2019) and Fatigue(Krigolson et al, 2021; Ruyi et al, 2017). All these findings show that MUSE can measure to some extend both bottom-up and top-down processes of the brain.…”

“…In terms of recored brain signal quality, MUSE has been shown to be able to capture Event Related Potentials (ERPs) (Krigolson et al, 2021; Krigolson et al, 2017). It has also been shown to be able various aspects of induced and intrinsic brain activity in various aspects of human behavior, such Stress (Asif et al, 2019; Phutela et al, 2022), Attention(Vortmann et al, 2022), Meditation (Kim et al, 2022; Sharma et al, 2022), Mindfulness (Hawley et al, 2021; Hunkin et al, 2021), Rapid diagnosis of stroke (Wilkinson et al, 2020), Drowsiness (LaRocco et al, 2020), Emotion Classification (Raheel et al, 2019) and Fatigue(Krigolson et al, 2021; Ruyi et al, 2017). In a recent study MUSE headsets were used in a hyper-scanning study in an art installation (Chen et al, 2022), which was previously using EMOTIV EPOC headsets (Dikker et al, 2021).…”

Classifying human emotions in HRI: applying global optimization model to EEG brain signals

“…The classifiers selected in this study were: SVM with four different kernels (Linear, Polynomial, RBF, and Sigmoid), KNN, Decision Tree (DT), RF, and Multi-Layer perceptrons (MLP). These models have previously shown excellent results in stress and emotion classification from EEG signals (Khosrowabadi et al, 2011 ; Wu et al, 2017 ; Al shargie et al, 2018 ; Nishtha et al, 2022 ; Rajendran et al, 2022 ).…”

A scalable and robust system for audience EEG recordings

“…In terms of recorded brain signal quality, MUSE has been shown to be able to capture Event Related Potentials (ERPs) [ 10 , 17 , 18 ]. It has also been shown to be able to capture various aspects of induced and intrinsic brain activity in various aspects of human behavior, such as Stress [ 19 , 20 ], Attention [ 21 ], Meditation [ 22 , 23 ], Mindfulness [ 24 , 25 ], Rapid diagnosis of stroke [ 26 ], Drowsiness [ 9 ], Emotion Classification [ 27 ] and Fatigue [ 17 , 28 ]. In a recent study MUSE headsets were used in a hyper-scanning study in an art installation [ 29 ], which was previously using EMOTIV EPOC headsets [ 14 ].…”