Human seizure detection using quadratic Rényi entropy

Abstract

In this study, the quadratic Rényi entropy is applied for seizure detection from human electroencephalography (EEG) signals. Quadratic Rényi entropy was combined with two different methods; the empirical mode decomposition (EMD) and discrete wavelet transform (DWT). The use of these two methods is justified since EEGs are non-linear and non-stationary signals. First, the EEG signal is decomposed into sub-signals using the EMD method or the DWT. Then, the quadratic Rényi entropy is used as an input feature. The k-nearest neighbor (k-NN) classifier algorithm extracted the features with 99.5%–100% accuracy.