Using Empirical Mode Decomposition with Spatio-Temporal dynamics to classify single-trial Motor Imagery in BCI

Area of Science:

• Signal Processing
• Biomedical Engineering
• Neuroscience

Background:

• Empirical Mode Decomposition (EMD) is a signal processing technique.
• Current methods often process signal sources in isolation.
• Brain-computer interfaces (BCIs) require effective electroencephalogram (EEG) analysis.

Purpose of the Study:

• Introduce a novel Spatio-Temporal Multivariate Empirical Mode Decomposition (ST-MEMD) method.
• Evaluate ST-MEMD's performance against traditional EMD for EEG data.
• Assess the impact of incorporating spatial and temporal information simultaneously.

Main Methods:

• Developed ST-MEMD, an extension of EMD considering spatial and temporal data.
• Applied ST-MEMD and standard EMD to single-trial EEG data.
• Utilized a Motor Imagery task for a two-class BCI problem.

Main Results:

• ST-MEMD demonstrated improved sensitivity and specificity due to spatial data integration.
• The addition of temporal data in ST-MEMD did not yield significant performance improvements.
• Both methods were tested on electroencephalogram data for brain-computer interfacing.

Conclusions:

• ST-MEMD offers potential for enhanced EEG analysis in BCIs through spatial data.
• Further research may be needed to optimize the temporal data processing component of ST-MEMD.
• The study highlights the importance of considering multi-dimensional data in signal processing for neuroscience applications.