EEGEpochNet: Self-supervised contrastive learning for automated EEG epoch rejection with multi-level feature construction

Key Points

• The aim is to develop a framework for automated EEG epoch rejection using self-supervised learning techniques.
• Implemented self-supervised contrastive learning approaches
• Constructed multi-level features from EEG data
• Evaluated the framework's reliability in EEG analysis
• Demonstrated parameter efficiency in EEG epoch rejection
• Established a basis for clinical-grade automation in EEG analysis
• Improved reliability in processing EEG data