Detecting weak periodic signals in EEG time series

Abstract

Recently many attempts had been devoted to the detection of weak periodic signals embedded in noisy chaotic signals using simulation methods and Duffing Oscillator System. In this paper, weak periodic signals in electroencephalogram (EEG) time series are investigated using a Duffing Equation based method. The EEG signals are collected from FR1-F3 and FR2-F4 channels in the scalp of an epileptic patient. In order to excite the Duffing equation by outer signals, the data is scaled down to a compatible range with the equation by multiplying it with a constant. After running the data with the Duffing Equation, weak periodic signals are obtained. As a result, the presence of weak periodic signals is observed in EEG time series. The amplitude and frequency values of the found weak periodic signals are calculated. Presence of weak periodic signals within the EEG signals can give us insight about the dynamic structure of the brain and functional brain activity.