ICA-Based EEG Classification Using Fuzzy C-mean Algorithm

In this paper we used Independent Component Analysis (ICA) model of EEG signals for preprocessing and then Discrete Wavelet Transform (DWT) analysis for feature extraction from EEG signal which this features are useful in BCI application. Then we used Fuzzy C-means (FCM) algorithm for recognition of some diseases like epileptic seizure, Cerebral Palsy (CP), etc. This project can be divided in three parts. The first part is EEG signal preprocessing using ICA. The second part is the feature extraction of normal and abnormal EEG using feature vectors derived from the wavelet analysis. The third part is the classification of normal and abnormal signals using FCM algorithm. In section II we explain EEG signal preprocessing, in section III we describe feature extraction and in section IV we explain the classification