A Feature Extraction Algorithm of Brain Network of Motor Imagination Based on a Directed Transfer Function.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Electroencephalography (EEG) is crucial for understanding brain activity.
• Accurate feature extraction from EEG signals is vital for brain-computer interfaces.
• Distinguishing between left- and right-hand motor imagery EEG signals presents a challenge.

Purpose of the Study:

• To propose and validate a novel multichannel correlation analysis method for EEG feature extraction.
• To enhance the classification accuracy of left- and right-hand motor imagery EEG signals.
• To evaluate the effectiveness of incorporating network information flow features.

Main Methods:

• Utilized Directed Transfer Function (DTF) to identify EEG channel connectivity and construct brain networks.
• Extracted network information flow features using DTF.
• Combined DTF features with traditional Autoregressive (AR) model parameters for an extended feature set.
• Employed Support Vector Machine (SVM) for signal classification.
• Compared classification performance across different channel counts (2, 10, 32).

Main Results:

• The enlarged feature set significantly improved classification accuracy compared to the traditional AR feature set.
• The multichannel analysis method proved more effective across different channel configurations.
• DTF-extracted network information flow features achieved higher classification performance than AR model features alone.
• The study verified the effectiveness of the proposed multichannel correlation analysis method.

Conclusions:

• The proposed multichannel correlation analysis method, incorporating DTF-derived network information flow features, enhances EEG signal classification for motor imagery.
• This extended feature set offers a more robust approach for distinguishing between left- and right-hand movements.
• The findings underscore the value of network analysis in advancing EEG-based brain-computer interfaces.