Search research articles

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

Early-life musical intervention programs neuroimmune activity and synaptic plasticity to enhance adult learning and memory.

Brain, behavior, and immunity·2026

Same author

Abnormal protein glycosylation in epileptogenesis: evidence from experimental models and clinical insights.

Reviews in the neurosciences·2026

Same author

A multivariate equivalence test based on Mahalanobis distance with a data-driven margin.

Journal of biopharmaceutical statistics·2026

Same author

Myoepithelioma-like tumor of the vulvar region with estrogen receptor negativity and <i>GSTT1::IGLC7</i> fusion: a case report.

Frontiers in oncology·2026

Same author

Intrinsic restoring enabled mode-selective frequency shifts in self-referenced humidity sensing.

Optics letters·2026

Same author

Mechanistic Insights into Active Components of Rosa Roxburghii Juice Against Fluoride-Induced Osteoarthritis.

Antioxidants (Basel, Switzerland)·2026

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

Search research articles

Related Experiment Video

Updated: Mar 8, 2026

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

Low-Rank Linear Dynamical Systems for Motor Imagery EEG.

Wenchang Zhang¹, Fuchun Sun², Chuanqi Tan²

¹The State Key Laboratory of Intelligent Technology and Systems, Computer Science and Technology School, Tsinghua University, FIT Building, Beijing 100084, China; Institute of Medical Equipment, Wandong Road, Hedong District, Tianjin, China.

Computational Intelligence and Neuroscience

|January 19, 2017

Summary

Linear dynamical systems (LDSs) offer a streamlined approach to motor imagery electroencephalography (MI-EEG) pattern recognition. This method simplifies feature extraction and classification, outperforming traditional spatiospectral techniques.

More Related Videos

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

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: Mar 8, 2026

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

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

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
Biomedical Engineering
Signal Processing

Background:

Motor imagery electroencephalography (MI-EEG) pattern recognition is crucial for brain-computer interfaces.
Common spatial pattern (CSP) and spatiospectral methods are effective but require extensive preprocessing.
These preprocessing steps can negatively impact classification accuracy.

Purpose of the Study:

To introduce linear dynamical systems (LDSs) as an efficient method for MI-EEG feature extraction and classification.
To develop a low-rank matrix decomposition approach to enhance the robustness of LDSs by reducing noise and resting state components.
To propose a low-rank LDS algorithm for improved feature subspace decomposition on finite Grassmannian.

Main Methods:

Utilized linear dynamical systems (LDSs) for simultaneous spatial and temporal feature matrix generation.
Implemented a low-rank matrix decomposition technique to denoise EEG signals and remove resting state components.
Developed a low-rank LDS algorithm for feature subspace decomposition on finite Grassmannian.

Main Results:

The proposed LDS-based methods achieved higher classification accuracies compared to conventional approaches like CSP and CSSP.
Experiments conducted on public datasets ('BCI Competition III Dataset IVa' and 'BCI Competition IV Database 2a') validated the effectiveness of the proposed methods.
The low-rank matrix decomposition significantly improved the robustness of the system.

Conclusions:

LDSs provide a cost-effective and efficient alternative for MI-EEG feature extraction and classification, eliminating the need for complex preprocessing.
The integration of low-rank matrix decomposition with LDSs enhances system robustness and performance.
The proposed low-rank LDS algorithm demonstrates superior performance in MI-EEG pattern recognition tasks.