Epileptic seizure classification in EEG signals using second-order difference plot of intrinsic mode functions.

Area of Science:

• Neurology
• Biomedical Signal Processing
• Machine Learning

Background:

• Epilepsy is a neurological disorder marked by unpredictable brain electrical disturbances.
• Electroencephalogram (EEG) is crucial for detecting epileptic seizures.
• Accurate classification of EEG signals is vital for epilepsy diagnosis and management.

Purpose of the Study:

• To develop and evaluate a new method for classifying ictal and seizure-free EEG signals.
• To utilize Empirical Mode Decomposition (EMD) and Second-Order Difference Plots (SODP) for feature extraction.
• To assess the effectiveness of Artificial Neural Network (ANN) classifier with the proposed features.

Main Methods:

• EEG signals were decomposed into Intrinsic Mode Functions (IMFs) using EMD.
• Second-Order Difference Plots (SODP) were generated from IMFs to create elliptical structures.
• The 95% confidence ellipse area from SODP of specific IMFs was used as a classification feature.
• An Artificial Neural Network (ANN) classifier was employed using features from the first and second IMFs.

Main Results:

• The proposed method successfully discriminated between seizure-free and epileptic seizure EEG signals.
• Feature space derived from the ellipse area parameters of the first and second IMFs yielded good classification performance.
• Experimental validation on the University of Bonn EEG database demonstrated the method's effectiveness.

Conclusions:

• The combination of EMD and SODP offers a promising approach for automated EEG-based epilepsy detection.
• The ellipse area feature extracted from IMFs is effective for discriminating between ictal and seizure-free states.
• This method holds potential for improving the accuracy and efficiency of epilepsy diagnosis.