Epileptic Seizure Classification using Battle Royale Search and Rescue optimization based Deep LSTM

Epilepsy is a major threat to society regarding the treatment time, cost, and unpredictable nature of the disease, imposing an urgent need for intelligent analysis. Electroencephalogram (EEG) is a commonly deployed test for detecting epilepsy that analyses the electrical activity of an individual's brain. In this work, an optimized deep sequential model is proposed to improve the seizure classification performance based on a hybrid feature set derived from EEG signals. A novel hybridized algorithm called Battle Royale Search and Rescue optimization (BRRO) is proposed for optimizing the deep learning model. Also, a proposed hybrid feature set utilizes the empirical mode decomposition (EMD), variational mode decomposition (VMD), and empirical wavelet transform (EWT). This feature set is created to capture the discriminative temporal property of the dataset. The proposed method is validated using publically available datasets. The results manifest that the proposed optimized algorithm provides better results in comparison to other existing alternatives.