Epileptic Seizure Recognition Utilizing Improved Chimp Optimization Algorithm with Deep Learning on EEG Signals

Epileptic seizure detection utilizing Electroencephalogram (EEG) signals is a significant application in medical diagnostics and healthcare. The EEG signals are electrical recordings of brain activity mainly utilized to monitor functions in the brain. Epileptic seizure detection in EEG aids in the analysis as well as management of epilepsy, a nervous disorder considered by existing seizures. Seizure detection using EEG signals is a very complex task that needs collaboration among medical specialists and Deep Learning (DL), Machine Learning (ML), and Neural Network (NN) experts to guarantee the reliability and accuracy of the recognition method for patients with epilepsy. DL methods, such as Convolutional NNs (CNNs) and Recurrent NNs (RNNs), are given training on labelled EEG repositories containing seizure and non-seizure parts. This article presents an Epileptic Seizure Recognition using an Improved Chimp Optimization Algorithm with DL (ESR-ICOADL) technique on EEG signals. The ESR-ICOADL technique aims to examine the EEG signals for detecting and classifying epileptic seizures. At a preliminary stage, the ESR-ICOADL technique applies the data preprocessing stage for converting the input data into valuable formats. For epileptic seizure recognition, the ESR-ICOADL technique applies a Bidirectional Gated Recurrent Unit (BiGRU) approach. Lastly, the hyperparameter tuning of the BiGRU approach could be boosted by utilizing ICOA, which supports accomplishing improved classification efficiency. The investigational analysis of the ESR-ICOADL approach is investigated on EEG datasets, and the simulated outputs illustrate the ESR-ICOADL model's significant results in diverse strategies.