APPLICATION OF FUZZY ARTMAP NEURAL NETWORKS FOR EPILEPTIC SPIKE DETECTION USING WAVELET FEATURE EXTRACTION

This paper aims to introduce two different classifier systems based on fuzzy ARTMAP neural network for the automatic detection of epileptic spikes in 19-channel human electroencephalogram these algorithm (EEG) are fast and delivers satisfactory results. EEG signals are decomposed into 4 sub-bands by means of Discrete Wavelet Transform (DWT). The inputs of the networks consist of two different features, which are extracted from the subbands 3 and 4. The performances of the classifiers introduced in this paper, are compared with each other’s and other similar systems, according to the sensitivity, specificity and selectivity values.