Enhancing EEG data quality and precision for cloud-based clinical applications: an evaluation of the SLOG framework.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Data Science

Background:

• Automation accelerates digital medicine delivery but requires robust data preprocessing.
• Electroencephalography (EEG) generates high-resolution neural data, susceptible to noise and artifacts.
• Ocular artifacts (EOG) are challenging to remove from EEG, often requiring manual inspection of Independent Component Analysis (ICA) results.

Purpose of the Study:

• To address the inaccuracy of automated Electroencephalography (EEG) artifact rejection.
• To introduce and validate a novel algorithm, Second Layer Inspection for EOG (SLOG), for improved EOG artifact removal.
• To enhance data fidelity and reduce computational costs in clinical EEG analysis.

Main Methods:

• Development of the Second Layer Inspection for EOG (SLOG) algorithm, utilizing spatial and temporal patterns of eye movements.
• Re-examination of Independent Component Analysis (ICA)-identified EOG artifacts to prevent accidental elimination of neural data.
• Validation of SLOG on both simulated and real-world EEG datasets.

Main Results:

• SLOG achieved a 99% precision rate on simulated datasets.
• SLOG demonstrated an 85% precision rate on real EEG datasets.
• The algorithm maintains data fidelity and precision, crucial for cloud-based applications and budget management.

Conclusions:

• Automated ICA for EOG artifact rejection in EEG is prone to inaccuracies.
• SLOG provides an effective auxiliary method for precise EOG artifact removal, improving data quality.
• Implementing SLOG enhances the efficiency and cost-effectiveness of EEG data processing in clinical settings.