Advanced Neural Network Optimization for Eeg-Based Epileptic Seizure Detection

Our study created a computer-assisted method that detects and classifies epileptic episodes using an optimized neural network. Big data methods and the exponential growth of healthcare data have enabled AI use in healthcare. AI helps healthcare providers reduce mistakes and retrieve vital data for fast decision-making. Radiology equipment using AI can provide correct data and eliminate the need for tissue samples. In order to attain the best degree of precision while working with large datasets, some techniques may degrade accuracy. Through a unique way, our team is improving performance and user acceptability with neural networks and deep learning. By adopting an optimized neural network, a computer-aided approach was devised to identify epileptic seizures using EEG data. SONN clustered input patterns into four categories. EEG signal feature extraction using DWT. This collected data was placed into a Genetic Algorithm-trained model that combines SONN and MLP methods. MLP networks were trained by clustering input samples using Self-organizing Neural Networks (SONN). The model classified samples as epileptic or non-epileptic using classifier. A comparison of models showed the recommended technique was effective. The hybrid method improved EEG signal categorization using SONN and MLP-GA.