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

Insight into the surface property modification-enhanced C<sub>3</sub>N<sub>4</sub> performance of photocatalytic nitrogen fixation.

Chemical communications (Cambridge, England)·2022
Same author

A matching algorithm with isotope distribution pattern in LC-MS based on support vector machine (SVM) learning model.

RSC advances·2022
Same author

Facile fabrication of long-chain alkyl functionalized ultrafine reduced graphene oxide nanocomposites for enhanced tribological performance.

RSC advances·2022
Same author

Efficient Metal-Oriented Electrodeposition of a Co-Based Metal-Organic Framework with Superior Capacitive Performance.

ChemSusChem·2022
Same author

Nanophotonic Approach to Study Excited-State Dynamics in Semiconductor Nanocrystals.

The journal of physical chemistry letters·2022
Same author

Benign Pancreatic Neurofibroma with Malignant Imaging Features: A Case Report and Literature Review.

Frontiers in surgery·2022

Related Experiment Video

Updated: Jul 17, 2025

Assessment and Communication for People with Disorders of Consciousness
07:37

Assessment and Communication for People with Disorders of Consciousness

Published on: August 1, 2017

9.1K

Towards best practice of interpreting deep learning models for EEG-based brain computer interfaces.

Jian Cui1, Liqiang Yuan2, Zhaoxiang Wang1

  • 1Research Center for Augmented Intelligence, Research Institute of Artificial Intelligence, Zhejiang Lab, Hangzhou, China.

Frontiers in Computational Neuroscience
|September 4, 2023
PubMed
Summary

Understanding deep learning models in electroencephalography-based brain-computer interfaces (EEG-BCI) is crucial. This study evaluates interpretation techniques, finding that method selection and sample-specific quality are key for reliable EEG-BCI model insights.

Keywords:
brain-computer interface (BCI)convolutional neural networkdeep learning interpretabilityelectroencephalography (EEG)layer-wise relevance propagation (LRP)

More Related Videos

A Single-Channel and Non-Invasive Wearable Brain-Computer Interface for Industry and Healthcare
06:34

A Single-Channel and Non-Invasive Wearable Brain-Computer Interface for Industry and Healthcare

Published on: July 7, 2023

2.4K
A Novel Experimental and Analytical Approach to the Multimodal Neural Decoding of Intent During Social Interaction in Freely-behaving Human Infants
11:14

A Novel Experimental and Analytical Approach to the Multimodal Neural Decoding of Intent During Social Interaction in Freely-behaving Human Infants

Published on: October 4, 2015

11.0K

Related Experiment Videos

Last Updated: Jul 17, 2025

Assessment and Communication for People with Disorders of Consciousness
07:37

Assessment and Communication for People with Disorders of Consciousness

Published on: August 1, 2017

9.1K
A Single-Channel and Non-Invasive Wearable Brain-Computer Interface for Industry and Healthcare
06:34

A Single-Channel and Non-Invasive Wearable Brain-Computer Interface for Industry and Healthcare

Published on: July 7, 2023

2.4K
A Novel Experimental and Analytical Approach to the Multimodal Neural Decoding of Intent During Social Interaction in Freely-behaving Human Infants
11:14

A Novel Experimental and Analytical Approach to the Multimodal Neural Decoding of Intent During Social Interaction in Freely-behaving Human Infants

Published on: October 4, 2015

11.0K

Area of Science:

  • Neuroscience
  • Machine Learning
  • Biomedical Engineering

Background:

  • Deep learning models achieve high performance in electroencephalography-based brain-computer interfaces (EEG-BCI).
  • Interpreting these models, often via heatmaps, is essential for understanding their decision-making processes.
  • The trustworthiness and accuracy of current interpretation methods for EEG-BCI remain unclear.

Purpose of the Study:

  • To quantitatively evaluate seven deep interpretation techniques for EEG-BCI models.
  • To assess the reliability and accuracy of these interpretation methods across various models and datasets.
  • To propose procedures for presenting trustworthy interpretation results in EEG-BCI.

Main Methods:

  • Quantitative evaluation of seven distinct deep interpretation techniques.
  • Testing across diverse deep learning models and EEG datasets.
  • Analysis of interpretation result quality concerning model structure and data types.

Main Results:

  • The choice of interpretation technique significantly impacts results.
  • Interpretation quality is inconsistent across individual samples, even with generally effective methods.
  • Model architecture and dataset characteristics influence interpretation accuracy.

Conclusions:

  • Selecting the appropriate interpretation technique is a critical first step for EEG-BCI.
  • Developing standardized procedures is necessary to ensure understandable and trustworthy model interpretations.
  • Proposed methods enhance the reliability of insights derived from EEG-BCI models.