Seizure prediction by analyzing EEG signal based on phase correlation.

Area of Science:

• Neurology
• Biomedical Engineering
• Signal Processing

Background:

• Epilepsy is a prevalent neurological disorder marked by recurrent seizures.
• Electroencephalogram (EEG) is crucial for diagnosing epileptic seizures.
• Predicting epileptic seizures from EEG signals presents challenges due to patient-specific signal variations.

Purpose of the Study:

• To develop a new feature extraction method for EEG signals to improve epileptic seizure prediction.
• To address the challenge of patient-specific variations in EEG data for seizure prediction.

Main Methods:

• Proposed a novel feature extraction approach utilizing phase correlation in EEG signals.
• Calculated relative changes between consecutive EEG signal segments.
• Combined signal changes with neighboring segments to extract predictive features.
• Classified preictal/ictal and interictal EEG signals using extracted features for seizure prediction.

Main Results:

• The proposed phase correlation method demonstrated a good prediction rate for epileptic seizures.
• Achieved greater consistency in prediction across different brain locations using benchmark datasets.
• Outperformed existing state-of-the-art methods in seizure prediction accuracy and consistency.

Conclusions:

• The novel phase correlation-based feature extraction method shows significant promise for accurate and consistent epileptic seizure prediction.
• This approach offers a potential advancement in the clinical application of EEG analysis for epilepsy management.