EEG-Based Emotion Recognition by Exploiting Fused Network Entropy Measures of Complex Networks across Subjects.

Longxin Yao,Xue Zhang,Yun Lu,Heng Li,Mingjiang Wang

doi:10.3390/e23080984

Abstract

It is well known that there may be significant individual differences in physiological signal patterns for emotional responses. Emotion recognition based on electroencephalogram (EEG) signals is still a challenging task in the context of developing an individual-independent recognition method. In our paper, from the perspective of spatial topology and temporal information of brain emotional patterns in an EEG, we exploit complex networks to characterize EEG signals to effectively extract EEG information for emotion recognition. First, we exploit visibility graphs to construct complex networks from EEG signals. Then, two kinds of network entropy measures (nodal degree entropy and clustering coefficient entropy) are calculated. By applying the AUC method, the effective features are input into the SVM classifier to perform emotion recognition across subjects. The experiment results showed that, for the EEG signals of 62 channels, the features of 18 channels selected by AUC were significant (p < 0.005). For the classification of positive and negative emotions, the average recognition rate was 87.26%; for the classification of positive, negative, and neutral emotions, the average recognition rate was 68.44%. Our method improves mean accuracy by an average of 2.28% compared with other existing methods. Our results fully demonstrate that a more accurate recognition of emotional EEG signals can be achieved relative to the available relevant studies, indicating that our method can provide more generalizability in practical use.

Highlights

After constructing the complex network corresponding to the EEG signals, for its comAfter constructing the complex network corresponding to the EEG signals, for its comprehensive characterization, we described the spatial characteristics of the brain network prehensive characterization, we described the spatial using characteristics theNDE
This study proposed a method for fusing network entropy measures, used to achieve effective emotion recognition results based on EEG signals across subjects
The main innovations of the fused network entropy measures method are as follows: (1) mapping the time series of EEG signals to complex networks using the visibility graph method; (2) exploiting the CCE and NDE features from the complex network, describing the spatial properties of EEG signals in the form of local and global information, respectively; (3) using the cross-subject emotion training method based on the SJTU Emotion EEG Dataset (SEED) dataset to overcome individual differences, improving the universality and generalizability of emotion recognition

Summary

Introduction

Current emotion recognition methods are mainly based on non-physiological signals and physiological signals. Physiological signals excellently reflect the functions of humans, with the advantages of objectivity and accuracy. Common physiological signals are diverse, and include electroencephalogram (EEG) [9], electromyogram (EMG) [10], and electrocardiogram (ECG) [11], etc. Among the above physiological signals, EEG signals can be obtained from the cerebral cortex using noninvasive devices, with the advantages of being direct, noninvasive and safe. Emotion states can be directly reflected by EEG signals related to corresponding brain regions. Based on the above case, emotion recognition based on EEG signals has attracted more and more attention due to its characteristics of being safe, noninvasive and intuitive [12,13,14,15]

Objectives

Methods

Results

Discussion

Conclusion