Visibility Graph from Adaptive Optimal Kernel Time-Frequency Representation for Classification of Epileptiform EEG.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Epileptic seizure detection from electroencephalogram (EEG) signals is a critical yet challenging problem in neurology.
• Accurate detection is vital for patient diagnosis, treatment, and management of epilepsy.

Purpose of the Study:

• To develop and validate a novel method for detecting epileptic seizures from EEG signals.
• To differentiate between healthy subjects and epilepsy patients, and between seizure-free and seizure states.

Main Methods:

• Utilized adaptive optimal kernel time-frequency representation to process EEG signals.
• Constructed complex networks from EEG data and analyzed topological properties (clustering coefficient, entropy, average degree).
• Combined network measures with energy deviation for classification and state distinction.

Main Results:

• The proposed method demonstrated high-accurate classification of epileptiform EEG signals.
• Successfully distinguished between healthy subjects and epilepsy patients.
• Effectively differentiated brain states during seizure-free intervals and epileptic seizures.

Conclusions:

• The novel method integrating time-frequency analysis and complex network theory offers a promising approach for accurate epileptic seizure detection.
• The findings suggest potential for improved diagnostic tools in epilepsy management.