Downloads: 99 | Views: 159 | Weekly Hits: ⮙2 | Monthly Hits: ⮙2

Research Paper | Biological Engineering | India | Volume 3 Issue 6, June 2014

Classification of Electroencephalogram Using Wavelet Transform and Neural Network

Sachin M. Rathod | S. C. Kulkarni

Abstract: Electroencephalogram (EEG) being a non-stationary signal its analysis using the Fourier Transform (FT) and Short Time Fourier Transform (STFT) is limited to a selection of window in which signal remains stationary. In this paper we will classify the Electroencephalogram (EEG) using Wavelet Transform (WT) and Feed Forward Neural Network. The database of Sleep EEG and Epileptic EEG are obtained from the www. physionet. org in the EDF (European Data File) file. The various features of EEG are obtained using WT these feature namely are delta; theta; alpha and beta. Any one of features described is dominant during a particular stages of sleep or when patient is alert; however with the age of patient; feature distribution with condition of patient may change i. e. dominant feature during particular stage of sleep is not same for different age group of patients. The feature vector obtained is given as input to Feed forward Neural Network (FFNN) for classification. Hence alertness level of a patient can be classified as sleep; drowsy and alert.

Keywords: EDF European Data File, EDFbrowser Electro-encephalogram, Feed Forward Neural Network, Fourier Transform FT, physioorgnet, Polyman, STFT, WT

Edition: Volume 3 Issue 6, June 2014,

Pages: 932 - 935