Classification of seizure and nonseizure EEG signals exploiting higher order statistics of the dominant Intrinsic mode function

In this paper, a method of seizure and non-seizure classification has been proposed based on the higher order statistics of the dominant Intrinsic mode function (IMF) resulting from the Empirical Mode Decomposition(EMD) of the EEG signals. Analyzing the temporal energy contents of different IMFs, it is found reasonable to determine the dominant IMF. In order to reduce the dimensionality, higher order statistics of the dominant IMF are employed to form the feature vector. The reduced feature vector thus formed is found effective for distinguishing seizure and non-seizure EEG signals when fed to a k-nearest neighborhood(k-NN) classifier. Extensive simulations are carried out using a benchmark EEG dataset. It is shown that the proposed method is capable of producing greater sensitivity, specificity, and accuracy in comparison to that obtained by using a state-of-the-art method employing the same EEG dataset and classifier.