Probability mapping based artifact detection and removal from single-channel EEG signals for brain-computer interface...

Area of Science:

• Neuroscience
• Signal Processing
• Biomedical Engineering

Background:

• Artifacts in electroencephalogram (EEG) signals significantly impair the performance of subsequent analysis algorithms, particularly in brain-computer interface (BCI) applications.
• Effective artifact detection and removal are crucial for reliable EEG-based decision-making and BCI classification.

Purpose of the Study:

• To develop and validate a novel approach for detecting and removing artifacts from single-channel EEG signals.
• To enhance the performance of BCI systems by improving the quality of processed EEG data.

Main Methods:

• A novel artifact detection method using four statistical measures: entropy, kurtosis, skewness, and periodic waveform index.
• Artifact removal employing stationary wavelet transform, guided by a user-defined probability threshold for artifactual epochs.

Main Results:

• The proposed method demonstrates superior performance in suppressing artifacts from both synthetic and real EEG data with minimal distortion.
• Evaluation on BCI datasets shows that artifact removal substantially improves BCI output for both event-related potential and motor-imagery tasks.

Conclusions:

• The developed algorithm offers a robust and efficient solution for EEG artifact removal.
• This work is expected to advance real-time EEG signal processing and contribute to the development of more accurate BCI applications.