Design Simulation and Assessment of EEG Based Expert System for Classification and Detection of Epileptic Seizure

Epilepsy is a neurological illness of around 1-2 percent world population prevalence. The feature of epilepsy is repeated seizures known as "epileptic seizures." The human brain is also the most outstanding and sophisticated organ of any system of human body. It has great space-time dynamics. The electroencephalographic signal is the spontaneous electrical activity recorded in the brain during a short period of time. The word EEG refers to a signal from the head that is obtained from the brain. The neuronal bombardment in the brain causes it. The EEG signal provides useful information on brain function and neurobiological illnesses, as it provides visual indications of the recorded waveform and enables approaches for computer-supporting signal processing to be characterized. This is why the most advanced digital signal processing techniques for EEG signal analysis are being used. Our research focuses on analyzing and classifying the signals received by EEG employing signals, such as waves. The EEG characteristics are taken from the statistical analysis of wavelet transformation. The second aim is to enhance classification accuracy once the feature has been extracted. A total of 300 EEG data subjects were analyzed. These data have been classified into three categories: normal, epileptic, epileptic and non-seizure. For this purpose, this research work has employed a back-propagation-based neural network classifier. The second aim is to enhance classification accuracy once the feature has been extracted. 100 subjects from the set and from the data dividing the proposed algorithmfor training, testing and validation were assessed for extraction and categorization of functions.