Artificial Neural Network for the Analysis of Electroencephalogram

Abstract

Electroencephalography is an important tool for diagnosing, monitoring and managing neurological disorders related to epilepsy. The presence of epileptiform activity in the electroencephalogram (EEG) confirms the diagnosis of epilepsy. During the seizures, the scalp of patients with epilepsy is characterized by high amplitude synchronized periodic EEG waveforms, reflecting abnormal discharge of a large group of neurons. Between the seizures, the electroencephalogram (EEG) of the patients who suffer from epilepsy is normally characterized by occasional spikes or spike and wave complexes (inter-ictal activity). It is difficult to detect these and sometimes is missed by the clinicians who observe the paper records. The purpose of the work describes the automated detection of epileptic events based on wavelet analysis of electroencephalogram. Three layered feedforward back-propagation artificial neural network (ANN) is designed to classify the epileptic seizure and non-epileptic seizure.