Related Experiment Video
Updated: Oct 4, 2025
08:45
Mapping Cortical Dynamics Using Simultaneous MEG/EEG and Anatomically-constrained Minimum-norm Estimates: an Auditory Attention Example
Published on: October 24, 2012
14.8K
Reducing calibration time in motor imagery-based BCIs by data alignment and empirical mode decomposition
1Dept. of Electronic Information Engineering, School of Information Engineering, Nanchang University, Nanchang, People's Republic of China.
Plos One
|February 8, 2022
Summary
This study introduces a new method to shorten brain-computer interface (BCI) calibration times. By generating artificial electroencephalogram (EEG) data, it reduces the need for extensive training data in motor imagery (MI) BCIs.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Brain-computer interfaces (BCIs) are limited by lengthy calibration periods.
- Motor imagery (MI) based BCIs require substantial training data for accurate classification.
Purpose of the Study:
- To develop a novel approach for reducing calibration time in MI-based BCIs.
- To augment training datasets by generating artificial electroencephalogram (EEG) data without compromising accuracy.
Main Methods:
- Empirical Mode Decomposition (EMD) and intrinsic mode function (IMF) mixing to generate artificial EEG data.
- Euclidean Alignment (EA) for aligning original training trials.
- Utilizing augmented datasets with Linear Discriminant Analysis (LDA) or Logistic Regression (LR) classifiers.
Main Results:
- The proposed algorithm significantly reduces the amount of training data required.
- Achieved comparable classification accuracy with reduced calibration time.
- Demonstrated effectiveness on two MI datasets.
Conclusions:
- The novel approach effectively reduces calibration time for MI-BCIs.
- Facilitates practical, real-world applications of BCI technology.
- Offers a promising solution for efficient BCI system development.
