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 Experiment Videos

Multiclass common spatial patterns and information theoretic feature extraction.

Moritz Grosse-Wentrup1, Martin Buss

  • 1Institute of Automatic Control Engineering (LSR), Technische Universität München, D-80290 München, Germany. moritzgw@ieee.org

IEEE Transactions on Bio-Medical Engineering
|July 18, 2008
PubMed
Summary
This summary is machine-generated.

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

A Conversational Brain-Artificial Intelligence Interface.

IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society·2026
Same author

Neuro-cognitive multilevel causal modeling: A framework that bridges the explanatory gap between neuronal activity and cognition.

PLoS computational biology·2024
Same author

Benchmarking brain-computer interface algorithms: Riemannian approaches vs convolutional neural networks.

Journal of neural engineering·2024
Same author

Predicting motor behavior: an efficient EEG signal processing pipeline to detect brain states with potential therapeutic relevance for VR-based neurorehabilitation.

Virtual reality·2023
Same author

Workshops of the Eighth International Brain-Computer Interface Meeting: BCIs: The Next Frontier.

Brain computer interfaces (Abingdon, England)·2023
Same author

Learning a Low-Dimensional Representation of a Safe Region for Safe Reinforcement Learning on Dynamical Systems.

IEEE transactions on neural networks and learning systems·2021
Same journal

Magnetic Resonance Spectroscopy Deep Learning with Magnetic Resonance Background Generator Enables In Vivo Metabolite Quantification of Hepatic Encephalopathy.

IEEE transactions on bio-medical engineering·2026
Same journal

Use of RPNIs and Implanted Electrodes for Prosthetic Wrist and Multi-Grip Hand Control during Functional Tasks: A Case Study.

IEEE transactions on bio-medical engineering·2026
Same journal

Healthy Limb Driven Prediction for Real Time Control of Unilateral Exoskeletons in Gait Rehabilitation.

IEEE transactions on bio-medical engineering·2026
Same journal

A Miniature Wearable Ultrasound System for Continuous Bladder Monitoring with Sleeping-Position-Robust Modeling Strategies.

IEEE transactions on bio-medical engineering·2026
Same journal

A Bi-objective Array Optimization Framework for Magnetocardiographic Source Imaging.

IEEE transactions on bio-medical engineering·2026
Same journal

A Dynamic Mutual Information Measure of Phase-Amplitude Coupling with Uncertainty Quantification.

IEEE transactions on bio-medical engineering·2026
See all related articles

This study enhances the Common Spatial Patterns (CSP) algorithm for brain-computer interfaces (BCIs) using information theory. The improved method boosts classification accuracy for electroencephalography/magnetoencephalography (EEG/MEG) data.

Area of Science:

  • Neuroscience
  • Biomedical Engineering
  • Machine Learning

Background:

  • Common Spatial Patterns (CSP) is widely used for spatial filtering in Brain-Computer Interfaces (BCIs).
  • Existing CSP methods have limitations regarding optimality and multiclass extensions.
  • Current multiclass CSP approaches often rely on heuristics.

Purpose of the Study:

  • To address the shortcomings of the CSP algorithm in BCI applications.
  • To establish a theoretical link between CSP and minimal classification error.
  • To develop an improved, information-theoretic approach for multiclass CSP.

Main Methods:

  • Utilized information theoretic feature extraction (ITFE) to analyze CSP.
  • Demonstrated CSP's maximization of mutual information approximation for two-class problems.

Related Experiment Videos

  • Proposed a novel method for multiclass CSP based on Joint Approximate Diagonalization (JAD) and Independent Component Analysis (ICA).
  • Developed a component selection strategy to maximize mutual information with class labels.

    • Main Results:

    • Established that CSP maximizes an approximation of mutual information for two-class BCIs.
    • Showed that multiclass CSP via JAD is equivalent to ICA.
    • The proposed method significantly improved classification accuracy by 23.4% on average for multiclass BCI data compared to standard multiclass CSP.
    • Eliminated the need for heuristics in multiclass CSP and allowed incorporation of prior class probabilities.

    Conclusions:

    • The information-theoretic framework provides a principled approach to CSP for BCI.
    • The proposed method offers a more optimal and robust solution for both two-class and multiclass BCI paradigms.
    • This advancement has the potential to enhance BCI performance and usability.