Epilepsy Seizure Detection Using Autoregressive Modelling and Multiple Layer Perceptron Neural Network

In this paper, we present a new method for epilepsy seizure detection based on autoregressive modelling. The method, termed linear prediction coding (LPC), is used to model ictal and seizure-free EEG signals. It is found that the modeling error energy is substantially higher for ictal EEG signals compared to seizure-free EEG signals. Moreover, it is known that ictal EEG signals have higher energy than seizure-free EEG signals. These two parameters are then given as inputs to train a Multiple Layer Perceptron (MLP). The trained MLP is then used to classify a set of EEG signals into ictal and seizure-free categories. It is found that the proposed method gives a classification accuracy of 94.67% when the MLP is trained with the Levenberg–Marquardt (LM) algorithm.