Hierarchical computer aided diagnostic system for seizure classification

EEG is the most economical and effective tool for understanding the complex dynamic behavior of the brain and studying its physiological states. In the present work, hierarchical computer aided diagnostic system (HCAD) for classification of normal, ictal and inter-ictal of EEG signals is proposed. In the present work, three different HCAD systems comprising of SVM, KNN and PNN classifiers are proposed. It is observed that the SVM based CAD system results in highest classification accuracy of 96% in comparison with 94% and 93.3% as obtained from KNN and PNN based HCAD systems. The promising results obtained from the present work indicate that the proposed SVM based HCAD system can be routinely used for seizure classification in clinical practice.