Analyzing spike I wave electroencephalogram signals for epilepsy based on Hilbert-Huang transformation

Abstract

In this paper, we use the Hilbert-Huang transform (HHT) analysis method to explore the time-frequency characteristics of spike waves for epilepsy symptoms. We obtained a sample of spike I wave and non-spike waves for HHT decomposition using a number of intrinsic mode functions of the Hilbert transform (HT) to determine the instantaneous spectrum, marginal spectrum, and Hilbert energy spectrum. We decomposed a number of intrinsic mode functions, the instantaneous spectrum and the Hilbert energy spectrum and compared the differences between spike and non-spike waves. The analysis results showed that the ratios of the normal wave to the referred total energy for IMF1, IMF2, and the residual function exceeded 10%. Furthermore, the ratios of the energy of the energy for IMF1, IMF2, IMF3 and the residual function of Spike I to their total energy also exceeded 10%. The ratios of the energy of IMF3 in the d band to its referred total energy for the EEG signal without spike wave, and of Spike I waves were 4.72%, and 6.75%, respectively. The ratios of the energy of IMF3 in the d band to its referred total energy for the EEG signal without spike wave is lower than that of the energy of IMF3 in the ν band to its referred total energy for the spike I wave.