Comparing EEG-Based Epilepsy Diagnosis Using Neural Networks and Wavelet Transform

Epilepsy is a common neurological disorder characterized by the recurrence of seizures, which can significantly impact lives patients. Electroencephalography (EEG) provide important physiological information on human brain activity be useful to diagnose epilepsy. However, manual analysis and visual inspection many EEG signals time-consuming may lead contradictory diagnoses doctors. play an role in diagnosis epilepsy, as quantification cerebral signal anomalies indicate condition pathology signal. In this study, we attempted develop two-step process for automated epilepsy using signals. first step, applied low-pass filter designed three intermediate filters different frequency bands, employing multi-layer neural networks. second used wavelet transform method data. The characteristics local are distribution epileptic model across whole surface. We also evaluated use two classifiers, artificial network (ANN) support vector machine (SVM), These classifiers were trained normal data able accurately distinguish between well other conditions. found that did not affect classification performance but provided better precision. developed process; incorporating filters, networks, led accurate efficient results paper show high accuracy rates both (92.38%) (95.5%) classifiers. Moreover, study highlighted effectiveness utilizing improved findings contribute ongoing efforts developing methods diagnosis, offering potential faster more reliable detection techniques enhance patient care outcomes.