Mental Task Classification Scheme Utilizing Correlation Coefficient Extracted from Interchannel Intrinsic Mode...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Brain computer interface (BCI) applications are rapidly expanding.
• Efficient classification of electroencephalogram (EEG) signals is crucial for BCI.
• Current feature extraction methods require improvement for diverse mental tasks.

Purpose of the Study:

• To develop an efficient feature extraction scheme for EEG-based mental task classification.
• To leverage interchannel EEG relationships for enhanced feature distinctiveness.
• To improve the accuracy of BCI systems through advanced signal processing.

Main Methods:

• Utilized empirical mode decomposition (EMD) to decompose EEG signals into intrinsic mode functions (IMFs).
• Extracted correlation coefficients from interchannel IMF data.
• Incorporated statistical features from intrachannel IMF data.
• Formed a feature matrix combining interchannel and intrachannel features.
• Employed support vector machine (SVM) classifiers with various kernels for classification.

Main Results:

• Achieved a very high classification accuracy using the proposed feature extraction scheme.
• Demonstrated the effectiveness of interchannel IMF correlations for distinguishing mental tasks.
• Outperformed existing methods in EEG signal classification accuracy.
• Validated the approach on an EEG dataset with five distinct mental tasks.

Conclusions:

• The proposed method effectively utilizes interchannel EEG relationships for robust feature extraction.
• This approach significantly enhances the accuracy of mental task classification in BCI systems.
• The combination of EMD, interchannel correlations, and SVM offers a promising direction for BCI research.