Abstract:During emotional changes, the brain generates many highly connected and highly concentrated hub regions. Thus, barely studying the whole-brain network architecture while ignoring the connection information between small-scale structures, will not be able to accurately differentiate the information interaction patterns of the brain under different emotions. The connections between high-degree nodes in the brain are stronger than those between low-degree nodes. The rich-club structure composed by those high-degree nodes is a hot topic in the current research of emotion recognition. Therefore, this paper investigates the dynamics of the rich-club structure generated by high-degree nodes in different time periods during different emotional changes. Firstly, the dynamic PLV brain network is constructed using non-overlapping time windows. Afterwards, the rich-club coefficients in the brain network are calculated and the nodes with higher degrees are selected as rich-club nodes within the coefficient range. Furthermore, the ReliefF algorithm is used to filter the frequency domain features of rich-club nodes and PLV-rich-club graph-theoretical features to derive the most emotionally relevant features for emotion recognition. The experimental results show that the composition of the rich-club is roughly the same but there are subtle differences over time. The smaller the valence, the larger the K-degree range, in which the structure of the rich-club is more stable, and the information interaction is more complex. The larger the valence, in which the information contained by the rich-club structure is more modular, and its ability to transmit information is stronger. It is revealed that the key to distinguishing different emotions is the brain information represented by the connections between small-scale structures in complex brain networks. The LOSO validation method was used for verification on the DEAP and SEED datasets, and the features based on the rich-club structure achieved 86.11% and 87.92% accuracy in the valence dimension, respectively.

Phase synchronization analysis of emotional EEG based on complex network theory

Identification of emotion associated brain functional network with phase locking value

Exploring EEG Emotion Recognition Through Complex Networks: Insights from the Visibility Graph of Ordinal Patterns

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

EEG decoding for musical emotion with functional connectivity features

Decoding emotion with phase–amplitude fusion features of EEG functional connectivity network

Construction and Analysis of Functional Brain Network Based on Emotional Electroencephalogram

Identifying Functional Brain Connectivity Patterns for EEG-Based Emotion Recognition.

Phase Synchronization Cluster Analysis for Brain Functional Connectivity

Analysis of a phase synchronized functional network based on the rhythm of brain activities

Investigating EEG-based functional connectivity patterns for multimodal emotion recognition

Topological EEG Nonlinear Dynamics Analysis for Emotion Recognition

EEG emotion recognition based on PLV-rich-club dynamic brain function network

Effect of Phase Clustering Bias on Phase-Amplitude Coupling for Emotional EEG

High Gamma Band EEG Closely Related to Emotion: Evidence from Functional Network

A study on the combination of functional connection features and Riemannian manifold in EEG emotion recognition

EEG-Based Functional Connectivity Representation using Phase Locking Value for Brain Network Based Applications

Emotion recognition based on group phase locking value using convolutional neural network

Network Analysis of Magnetoencephalogram Signals in Schizophrenia Patients When Viewing Emotional Facial Stimuli

The part and the whole: how single nodes contribute to large-scale phase-locking in functional EEG networks

Abnormal Brain Function Network Analysis Based on EEG and Machine Learning