Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

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

Amyloid PET Quantitation and Centiloid Thresholds in the Diagnosis of Alzheimer Disease: An Individual Participant Data Meta-Analysis.

JAMA·2026
Same author

SEM mapping of sequence-specific protein-DNA interactions on long DNA molecules.

Nucleic acids research·2026
Same author

The feasibility of [<sup>18</sup>F]PSMA-1007 PET/CT for response assessment in prostate cancer following neoadjuvant androgen deprivation therapy compared with mpMRI.

Cancer imaging : the official publication of the International Cancer Imaging Society·2026
Same author

Modal contrast engineering in ultraviolet and visible metalenses enabled by material-selective hybridization.

Nature communications·2026
Same author

Genetic and pharmacological inactivation of peptidoglycan remodeling increases antibiotic susceptibility of vancomycin-resistant Enterococcus faecium.

Nature communications·2026
Same author

Comparing machine and deep learning models for pediatric anxiety classification using structured EHRs and area-based measures of health data.

PloS one·2026

Related Experiment Video

Updated: Jul 11, 2025

Author Spotlight: Enhancing Neurorehabilitation Through EEG, Motor Imagery, and Virtual Reality
10:14

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

Published on: May 10, 2024

1.0K

Spatio-Temporal Explanation of 3D-EEGNet for Motor Imagery EEG Classification Using Permutation and Saliency.

Donghyun Park, Hoonseok Park, Sangyeon Kim

    IEEE Transactions on Neural Systems and Rehabilitation Engineering : a Publication of the IEEE Engineering in Medicine and Biology Society
    |November 7, 2023
    PubMed
    Summary

    This study introduces 3D-EEGNet, a 3D convolutional neural network (CNN) model that enhances both performance and explainability for motor imagery (MI) brain-computer interfaces (BCI) using electroencephalogram (EEG) data.

    More Related Videos

    Mapping Cortical Dynamics Using Simultaneous MEG/EEG and Anatomically-constrained Minimum-norm Estimates: an Auditory Attention Example
    08:45

    Mapping Cortical Dynamics Using Simultaneous MEG/EEG and Anatomically-constrained Minimum-norm Estimates: an Auditory Attention Example

    Published on: October 24, 2012

    14.7K
    Concurrent EEG and Functional MRI Recording and Integration Analysis for Dynamic Cortical Activity Imaging
    11:28

    Concurrent EEG and Functional MRI Recording and Integration Analysis for Dynamic Cortical Activity Imaging

    Published on: June 30, 2018

    11.7K

    Related Experiment Videos

    Last Updated: Jul 11, 2025

    Author Spotlight: Enhancing Neurorehabilitation Through EEG, Motor Imagery, and Virtual Reality
    10:14

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

    Published on: May 10, 2024

    1.0K
    Mapping Cortical Dynamics Using Simultaneous MEG/EEG and Anatomically-constrained Minimum-norm Estimates: an Auditory Attention Example
    08:45

    Mapping Cortical Dynamics Using Simultaneous MEG/EEG and Anatomically-constrained Minimum-norm Estimates: an Auditory Attention Example

    Published on: October 24, 2012

    14.7K
    Concurrent EEG and Functional MRI Recording and Integration Analysis for Dynamic Cortical Activity Imaging
    11:28

    Concurrent EEG and Functional MRI Recording and Integration Analysis for Dynamic Cortical Activity Imaging

    Published on: June 30, 2018

    11.7K

    Area of Science:

    • Neuroscience
    • Artificial Intelligence
    • Biomedical Engineering

    Background:

    • Convolutional neural network (CNN) models show promise for motor imagery (MI) brain-computer interfaces (BCI) using electroencephalogram (EEG).
    • Interpreting CNN models for EEG-based BCI classification remains a significant challenge.
    • Existing explainable artificial intelligence (XAI) techniques have limitations in effectively explaining these complex models.

    Purpose of the Study:

    • To propose 3D-EEGNet, a 3D CNN model, to enhance both the performance and explainability of MI EEG classification.
    • To compare the performance of 3D-EEGNet against existing 2D CNN models like EEGNet.
    • To develop and evaluate novel XAI methods for spatio-temporal explanation of 3D-EEGNet.

    Main Methods:

    • Development of 3D-EEGNet, a 3D convolutional neural network architecture.
    • Application of a permutation-based XAI method for initial model explanation.
    • Design of a novel technique using normalized discounted cumulative gain (NDCG) to select efficient saliency-based XAI methods.
    • Evaluation of DeepLIFT as a fast spatio-temporal explanation method.

    Main Results:

    • 3D-EEGNet demonstrated improved MI classification accuracies on two datasets compared to the 2D EEGNet, with average improvements of 1.8% and 6.1%.
    • The permutation-based XAI method provided reliable explanations for 3D-EEGNet.
    • The NDCG-based selection identified DeepLIFT as a fast and effective XAI method, showing results similar to the permutation-based approach.
    • Fast spatio-temporal explanations using DeepLIFT offered deeper insights into classification results and key MI EEG properties.

    Conclusions:

    • 3D-EEGNet offers a significant advancement in both performance and explainability for MI EEG classification.
    • The developed NDCG-based selection and DeepLIFT provide a practical approach for fast and reliable spatio-temporal explanations of CNN models in BCI.
    • This research contributes to a better understanding of MI EEG signals and improves the interpretability of deep learning models in BCI applications.