Epileptic seizure detection in EEG signal with GModPCA and support vector machine

Area of Science:

• Neurology
• Signal Processing
• Machine Learning

Background:

• Epilepsy is a common neurological disorder characterized by recurrent seizures.
• Electroencephalograms (EEGs) are used to detect epilepsy, but manual inspection is time-consuming and prone to error.
• Automated seizure detection frameworks are crucial for accurate and efficient epilepsy diagnosis.

Purpose of the Study:

• To propose an effective feature extraction and classification framework for automated epileptic seizure detection in EEG signals.
• To evaluate the performance of Global modular Principal Component Analysis (GModPCA) for feature extraction and Support Vector Machine (SVM) for classification.
• To demonstrate the analytical advantages of GModPCA over traditional Principal Component Analysis (PCA) in terms of complexity.

Main Methods:

• Feature extraction was performed using GModPCA, a variation of PCA.
• Classification of seizure and non-seizure EEG signals was conducted using an SVM with a radial basis function kernel.
• The framework's performance was rigorously evaluated using a benchmark epilepsy EEG dataset and 10-fold cross-validation.

Main Results:

• The GModPCA and SVM framework achieved 100% accuracy in classifying normal and epileptic EEG signals.
• Experimental results demonstrated superior performance compared to existing methods in the literature.
• GModPCA was analytically proven to have lower time and space complexities than PCA.

Conclusions:

• GModPCA combined with SVM offers a highly accurate and efficient solution for automated epileptic seizure detection.
• The proposed framework significantly improves upon traditional visual inspection methods for EEG analysis.
• GModPCA presents a valuable advancement in feature extraction for neurological signal processing.