EPIC-NET: EEG-based epilepsy classification and brain localization using Optuna wave-gated recurrent unit network

Key Points

• The EPIC-NET model improves accuracy by 5.92, 10.02, and 0.59% over RNN, SVM, and CNN, respectively.
• The model utilizes an EEG-based approach for classification and localization of epilepsy types.
• Assessment using a wave-gated recurrent unit network enhances performance metrics significantly.
• Potential applications include better diagnostic accuracy for epilepsy, aiding treatment strategies.