Related Experiment Video
Updated: May 14, 2026
11:25
Simultaneous Scalp Electroencephalography (EEG), Electromyography (EMG), and Whole-body Segmental Inertial Recording for Multi-modal Neural Decoding
Published on: July 26, 2013
Combination of amplitude and phase features under a uniform framework with EMD in EEG-based Brain-Computer Interface
1Shenzhen Key Lab of Neuropsychiatric Modulation, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, China. weo.he@siat.ac.cn
Summary
This study enhances Brain-Computer Interface (BCI) classification by combining EEG amplitude and phase features using Empirical Mode Decomposition. The novel approach significantly improves accuracy for motor imagery tasks.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Brain-Computer Interface (BCI) systems rely on electroencephalography (EEG) signals to interpret user intentions.
- Amplitude and phase variations in EEG are crucial for decoding movement intentions and mental tasks.
- Integrating amplitude and phase information offers a more comprehensive feature set for BCI.
Purpose of the Study:
- To develop a unified feature extraction framework for BCI by combining EEG amplitude and phase information.
- To improve the classification accuracy of BCI systems for motor imagery and mental tasks.
- To investigate the online feasibility of the proposed integrated feature extraction method.
Main Methods:
- Utilized Common Spatial Pattern (CSP) for amplitude and Phase Locking Value (PLV) for phase extraction.
- Employed Empirical Mode Decomposition (EMD) as a filter bank for optimized band selection and precise phase calculation.
- Applied Sequential Floating Forward Selection (SFFS) to select the most discriminative features.
Main Results:
- The integrated method demonstrated a statistically significant average increase in classification accuracy of 5.4% compared to traditional CSP.
- Improvements were observed on both public (2.0%) and recorded (8.7%) datasets.
- Preliminary investigations indicated comparable performance for online implementation versus offline analysis.
Conclusions:
- Combining EEG amplitude and phase features via EMD offers a superior approach for BCI classification.
- The proposed method enhances the decoding of brain states for improved BCI performance.
- The technique shows promise for real-time BCI applications.
