Knowledge-based identification of sleep stages based on two forehead electroencephalogram channels

Area of Science:

• Neuroscience
• Biomedical Engineering
• Sleep Medicine

Background:

• Sleep quality assessment is crucial for managing sleep pathologies.
• Polysomnography (PSG) is the gold standard but is complex and inconvenient for home use.
• There is a need for practical, at-home sleep monitoring solutions.

Purpose of the Study:

• To develop and validate a novel sleep stage classification system using forehead electroencephalogram (EEG) signals.
• To assess the system's performance in classifying five sleep stages.
• To establish forehead EEG as a viable alternative for convenient home-based sleep quality evaluation.

Main Methods:

• Utilized two forehead EEG channels (FP1 and FP2) for signal acquisition.
• Developed a system to extract key sleep features from forehead EEG.
• Employed a relevance vector machine (RVM) for classifying five sleep stages.
• Conducted leave-one-subject-out cross-validation for performance assessment.

Main Results:

• The system achieved an average accuracy of 76.7 ± 4.0% in classifying five sleep stages.
• The average kappa score was 0.68 ± 0.06, indicating good agreement.
• Forehead EEG features identified were closely related to clinical expert knowledge.
• The system demonstrated robustness against individual differences.

Conclusions:

• The proposed forehead EEG-based system provides a practical and convenient method for at-home sleep stage classification.
• This approach offers a reliable alternative to traditional PSG for monitoring sleep quality.
• The system has the potential to improve public healthcare by facilitating accessible sleep assessments.