Epileptic Seizure Classification Using Battle Royale Search and Rescue Optimization-Based Deep LSTM.

Area of Science:

• Neurology
• Artificial Intelligence
• Signal Processing

Background:

• Epilepsy poses significant societal challenges due to its unpredictable nature and high treatment costs.
• Electroencephalogram (EEG) is a critical diagnostic tool for analyzing brain electrical activity in epilepsy detection.
• There is an urgent need for intelligent analysis methods to improve epilepsy diagnosis and management.

Purpose of the Study:

• To develop an optimized deep sequential model for enhanced seizure classification from EEG signals.
• To introduce a novel hybridized Battle Royale Search and Rescue optimization (BRRO) algorithm for deep learning (DL) model optimization.
• To create a hybrid feature set using advanced signal processing techniques for capturing temporal EEG data properties.

Main Methods:

• Utilized empirical mode decomposition, variational mode decomposition, and empirical wavelet transform for hybrid feature extraction from EEG signals.
• Developed a novel hybridized Battle Royale Search and Rescue optimization (BRRO) algorithm to optimize a deep learning (DL) model.
• Implemented and validated an optimized deep sequential model for seizure classification using publicly available EEG datasets.

Main Results:

• The proposed optimized deep learning model demonstrated superior seizure classification performance compared to existing methods.
• The hybrid feature set effectively captured the complex temporal dynamics of EEG signals.
• The BRRO algorithm successfully optimized the DL model, leading to improved accuracy.

Conclusions:

• The developed optimized deep sequential model offers a promising advancement in intelligent epilepsy analysis.
• The novel BRRO algorithm provides an effective method for optimizing deep learning models in medical applications.
• This research contributes to more accurate and efficient epilepsy diagnosis through advanced EEG signal processing and AI.