Artifact processing in computerized analysis of sleep EEG - a review.

Area of Science:

• Neuroscience
• Sleep Medicine
• Biomedical Engineering

Background:

• Quantitative analysis of sleep electroencephalogram (EEG) data offers significant insights into sleep research.
• Data contamination by artifacts can lead to inaccurate and misleading research findings.
• Reliable artifact processing is crucial for developing automated sleep analysis systems.

Purpose of the Study:

• To review and outline a comprehensive artifact processing strategy for sleep EEG data.
• To emphasize the importance of artifact management in achieving valid quantitative sleep analysis.
• To describe methods for minimizing and identifying artifacts in sleep EEG recordings.

Main Methods:

• Implementing high-quality recording techniques to reduce avoidable technical artifacts.
• Employing artifact minimization procedures to correct for interference (e.g., ocular, ECG) and preserve data.
• Utilizing artifact identification procedures to detect and exclude epochs with remaining artifacts (e.g., movement, muscle).

Main Results:

• A multi-step strategy for processing sleep EEG artifacts is proposed.
• The strategy aims to improve the accuracy of quantitative sleep analysis.
• Review includes descriptions of artifact types and examples across sleep stages.

Conclusions:

• A robust artifact processing strategy is fundamental for accurate automated sleep analysis.
• Minimizing, correcting, and identifying artifacts ensures the integrity of sleep EEG data.
• This approach enhances the reliability and validity of quantitative sleep research findings.