Discrete wavelet transform based seizure detection in newborns EEG signals

This paper proposes a novel method for detecting newborns seizure events from electroencephalogram (EEG) data. The detection scheme is based on the discrete wavelet transform (DWT) of the EEG signals. The number of zero-crossings, the average distance between adjacent zero-crossings, the number of extrema, and the average distance between adjacent extrema of the wavelet coefficients (WCs) of certain scales are extracted to form a feature set. The extracted feature set is then fed to an artificial neural network (ANN) classifier to organize the EEG signals into seizure and non- seizure activities. In this study, the training and test sets were obtained from EEG data acquired from 1 and 5 other neonates, respectively, with ages ranging from 2 days to 2 weeks. The obtained results show that on the average 95% of the EEG seizures were detected by the proposed scheme.