METHOD AND ALGORITHM FOR DIAGNOSTIC OF EPILEPTIC EEG SIGNALS USING THE ADAPTIVE ORTHOGONAL TRANSFORM

摘要

Electroencephalography (EEG) is one of the most used techniques for evaluating the functional status of the brain. It is essential for diseases’ diagnosis such as epilepsy. This pathology results from a cerebral dysfunction. The diagnosis of this pathology consists of detecting the appearance of paroxysmal activities in the EEG signals. The diagnostic of Epilepsy in EEG plays a crucial role in Computer Aided Diagnosis system (CAD). In this article, we suggest an approach based on the orthogonal adaptive transformation theory which makes it possible to extract the informative features of the EEG signals. The size of the vectors of the informative features obtained by this method is very short. This will allow to improve the quality of signals analysis and to increase their certainty of diagnosis