Entropy based features in FAWT framework for automated detection of epileptic seizure EEG signals

Flexible analytic wavelet transform (FAWT) is suitable for the study of oscillatory signals like electroencephalogram (EEG) signals with versatile features such as shift in-variance, tunable oscillatory properties and flexible time-frequency domain. In this paper, we propose an automated method for the classification of seizure and non-seizure EEG signals using FAWT and entropy-based features such as Stein's unbiased risk estimator (SURE) entropy, log energy entropy, and Shannon entropy. The obtained features are given as input to robust energy-based least squares twin support vector machines (RELS-TSVM) for classification. The proposed method has been implemented on publicly available epilepsy database (Bonn University EEG database) and is comparable with the existing methods with a maximum accuracy of 100% for the classification of seizure and non-seizure EEG signals.