Abstract
A big number of people all over the world is affected by epilepsy. Electroencephalography (EEG) is a crucial component in the evaluation of epilepsy. Onset seizure detection is essential to prevent the seizure activity and improve patients' life quality. Presurgical treatment, precise assessment, seizure prevention, and emergency alerts all depend on the quick detection of seizure onset. Moreover, manual examination of EEG signals is boring and time-consuming task. Numerous automated epileptic seizure detection schemes have been developed to help neurologists. This paper focus on the realization of an efficient adaptive rate solution for the epileptic seizure detection. The signal is respectively digitized and segmented with an event-driven ADC (EDADC) and an activity selection algorithm (ASA). The segments are uniformly resampled and conditioned. In next step, the autoregressive (AR) Burg modelling is employed to extract the features. Afterwards, the extracted features are utilized for classification. It is demonstrated that the suggested system-processing load is adjusted as a function of the incoming signal disparities. It allows the suggested solution to attain a remarkable reduction in the processing activity and consumption of power compared to the counter classical ones. The overall system classification precision is also compared with the counter classical one. It confirms that the prospect of using the suggested system for an effective automatic epileptic seizure detection.