Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Characteristics, Clinical Manifestations, and Severe Disease Risk of Bordetella pertussis Infection in Children: A tNGS-based Retrospective Study.

International journal of infectious diseases : IJID : official publication of the International Society for Infectious Diseases·2026

Same author

Microbial spectrum, co-detection patterns, and clinical correlations in 10,153 hospitalized children with acute respiratory infections: a large-scale tNGS analysis.

Journal of translational medicine·2026

Same author

Peroxisome-derived and liver-specific hydroxyacid oxidase 1-mediated hydrogen peroxide promotes ferroptosis by augmenting peroxisomal ROS.

Journal of advanced research·2026

Same author

Predictive value of early blood glucose trajectory for poor prognosis in ARDS patients.

International journal of medical informatics·2026

Same author

Molecular tuning of DNA framework-programmed silicification by cationic silica cluster attachment.

Proceedings of the National Academy of Sciences of the United States of America·2026

Same author

Targeted next-generation sequencing for pediatric lower respiratory tract infections: a retrospective study.

Frontiers in cellular and infection microbiology·2025

Search research articles

Related Experiment Video

Updated: Aug 11, 2025

Conscious and Non-conscious Representations of Emotional Faces in Asperger's Syndrome

Conscious and Non-conscious Representations of Emotional Faces in Asperger's Syndrome

Published on: July 31, 2016

EEG emotion recognition using improved graph neural network with channel selection.

Xuefen Lin¹, Jielin Chen¹, Weifeng Ma¹

¹School of Information and Electronic Engineering, Zhejiang University of Science and Technology, Hangzhou 310023, PR China.

Computer Methods and Programs in Biomedicine

|February 6, 2023

Summary

This summary is machine-generated.

This study introduces an improved graph convolution model for electroencephalography (EEG) emotion classification. The model achieves high accuracy even with reduced EEG channels, making affective computing more cost-effective.

Keywords:

Attention mechanism Convolutional neural network EEG classification Graph neural network

More Related Videos

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

Brain Mapping Using a Graphene Electrode Array

Brain Mapping Using a Graphene Electrode Array

Published on: October 20, 2023

Related Experiment Videos

Last Updated: Aug 11, 2025

Conscious and Non-conscious Representations of Emotional Faces in Asperger's Syndrome

Conscious and Non-conscious Representations of Emotional Faces in Asperger's Syndrome

Published on: July 31, 2016

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention

Published on: December 15, 2023

Brain Mapping Using a Graphene Electrode Array

Brain Mapping Using a Graphene Electrode Array

Published on: October 20, 2023

Area of Science:

Artificial Intelligence
Neuroscience
Machine Learning

Background:

Emotion classification using electroencephalography (EEG) is vital for AI applications in healthcare.
Variable EEG channel numbers in complex environments hinder accurate brain information transfer simulation.
Existing models face challenges in adapting to diverse EEG data acquisition settings.

Purpose of the Study:

To propose an improved graph convolution model for robust EEG-based emotion classification.
To address the challenge of variable EEG channel numbers in real-world applications.
To develop a dynamic channel selection mechanism for efficient affective computing.

Main Methods:

A hybrid model combining 1D and graph convolution for intra- and inter-channel EEG feature extraction.
Incorporation of functional connectivity into the graph structure to better represent brain region relationships.
An attention-based dynamic channel selection strategy for adjustable feature subset utilization.

Main Results:

Achieved high average accuracies of 90.74% (DEAP-Twente), 91% (DEAP-Geneva), and 90.22% (SEED).
Demonstrated strong performance with only 20% of EEG channels, reaching average accuracies of 82.78%, 84%, and 83.93% respectively.
Outperformed most existing models in emotion classification tasks across multiple datasets.

Conclusions:

The proposed model effectively performs emotion classification in complex EEG environments.
The dynamic channel selection method significantly reduces the cost of affective computing.
The model offers a promising solution for practical healthcare applications like autism research and prenatal emotion monitoring.