EEG Subband Analysis using Approximate Entropy for the Detection of Epilepsy

Abstract

Epilepsy is a neurological disorder which affects the nervous system. Epileptic seizures are due to sudden hyperactivity in certain parts of the brain. Electroencephalogram (EEG) is the commonly used modality for the detection of epilepsy. Automatic seizure detection helps in diagnosis and monitoring of epilepsy especially during long term recordings of EEG. This paper presents non linear analysis of EEG for the detection of epilepsy using approximate entropy (ApEn). The proposed method involves ApEn measured from EEG subbands applied as features to an artificial neural network (ANN) classifier. The ApEn measured from delta, theta, alpha, beta and gamma subbands of normal EEG, ictal and inter ictal EEGs are used as features. In the present work detection of epilepsy is considered as a two class problem. In the first case the classification is done between normal and ictal EEGs and in the second case, classification is done between normal and inter ictal EEGs. For both cases artificial neural networks with back propagation training are used as classifiers. The classification accuracy of 100% is obtained for normal and ictal groups and that of 98.9% is obtained for normal and inters ictal EEGs. Keywords: Electroencephalogram (EEG), ictal and inter ictal EEG, approximate entropy, neural network classifier.