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

Co-Creating the Cities of the Future.

Sensors (Basel, Switzerland)·2016

See all related articles

Search research articles

Related Experiment Video

Updated: Jul 15, 2025

Author Spotlight: Using Motor Imagery Brain-Computer Interface to Improve Motor and Cognitive Function in Stroke Patients

Author Spotlight: Using Motor Imagery Brain-Computer Interface to Improve Motor and Cognitive Function in Stroke Patients

Published on: September 1, 2023

Enhancing Cross-Subject Motor Imagery Classification in EEG-Based Brain-Computer Interfaces by Using Multi-Branch

Radia Rayan Chowdhury¹, Yar Muhammad^1,2, Usman Adeel¹

¹Department of Computing & Games, School of Computing, Engineering & Digital Technologies, Teesside University, Middlesbrough TS1 3BX, UK.

Sensors (Basel, Switzerland)

|September 28, 2023

Summary

A new multi-branch deep learning model, EEGNet Fusion V2, improves brain-computer interface (BCI) accuracy for classifying motor imagery from electroencephalogram (EEG) signals across different subjects. This advanced model shows superior performance on public datasets compared to existing methods.

Keywords:

brain–computer interface (BCI)convolutional neural network (CNN)deep learning electroencephalography (EEG)fusion network motor imagery (MI)

More Related Videos

Author Spotlight: Enhancing Neurorehabilitation Through EEG, Motor Imagery, and Virtual Reality

Author Spotlight: Enhancing Neurorehabilitation Through EEG, Motor Imagery, and Virtual Reality

Published on: May 10, 2024

Electroencephalography Network Indices as Biomarkers of Upper Limb Impairment in Chronic Stroke

Electroencephalography Network Indices as Biomarkers of Upper Limb Impairment in Chronic Stroke

Published on: July 14, 2023

Related Experiment Videos

Last Updated: Jul 15, 2025

Author Spotlight: Using Motor Imagery Brain-Computer Interface to Improve Motor and Cognitive Function in Stroke Patients

Author Spotlight: Using Motor Imagery Brain-Computer Interface to Improve Motor and Cognitive Function in Stroke Patients

Published on: September 1, 2023

Author Spotlight: Enhancing Neurorehabilitation Through EEG, Motor Imagery, and Virtual Reality

Author Spotlight: Enhancing Neurorehabilitation Through EEG, Motor Imagery, and Virtual Reality

Published on: May 10, 2024

Electroencephalography Network Indices as Biomarkers of Upper Limb Impairment in Chronic Stroke

Electroencephalography Network Indices as Biomarkers of Upper Limb Impairment in Chronic Stroke

Published on: July 14, 2023

Area of Science:

Neuroscience
Computer Science
Machine Learning

Background:

Brain-computer interfaces (BCIs) enable interaction via neural activity, typically using electroencephalogram (EEG) signals.
Classifying subject-independent motor imagery EEG data is challenging due to high individual variability.
Deep learning, particularly Convolutional Neural Networks (CNNs), shows promise for enhancing feature extraction and classification accuracy in BCIs.

Purpose of the Study:

To develop and evaluate a novel multi-branch 2D CNN model for subject-independent motor imagery classification using EEG data.
To assess the performance of the proposed model against established BCI deep learning architectures.
To analyze the accuracy and computational cost of the new model on public BCI datasets.

Main Methods:

A five-branch 2D CNN architecture (EEGNet Fusion V2) was designed with unique hyperparameters for each branch.
The model was trained and validated on three public EEG motor imagery datasets: eegmmidb, BCI IV-2a, and BCI IV-2b.
Performance was benchmarked against several existing models including EEGNet, ShallowConvNet, DeepConvNet, MMCNN, and EEGNet_Fusion.

Main Results:

EEGNet Fusion V2 achieved high cross-subject classification accuracies: 89.6% (actual) and 87.8% (imagined) on eegmmidb.
The model obtained scores of 74.3% and 84.1% on the BCI IV-2a and IV-2b datasets, respectively.
The proposed model outperformed the compared methods on the tested datasets, although with a higher computational cost (approx. 3.5x EEGNet_Fusion).

Conclusions:

The proposed multi-branch CNN, EEGNet Fusion V2, demonstrates significant improvements in subject-independent motor imagery classification for BCIs.
The model offers a promising advancement over existing deep learning approaches for EEG-based BCI applications.
Further research may focus on optimizing computational efficiency while maintaining high classification performance.