Neonates' EEG spike detection using a time-frequency approach

Abstract

This paper proposes a new time-frequency (TF) based approach for detecting spikes in newborns' EEG signals. Since the automated detection of spikes in the EEG is an important deal in the diagnosis of epilepsy and is a goal sought by many researchers. in the proposed method, first a TF representation of the EEG epoch is computed and then its average over all frequencies is used to distinguish between the EEG spikes and the background. The method is used in conjunction with a few time-frequency distributions (TFDs) to detect spikes in the EEG signals acquired from 5 neonates, with ages ranging from two days to two weeks. The obtained results show that the short-time Fourier transform (STFT) outperforms the other TFDs with the average detection rate of 96%.