AN ENHANCED METAHEURISTIC-BASED DEEP LEARNING MODEL FOR EPILEPTIC SEIZURE DETECTION AND PREDICTION

Epilepsy is a non-communicable neurological disorder that affects more than 50 million people worldwide. The disease is characterized by rapid and uncontrollable body movements. Given the astronomical number of patients diagnosed with this disease and the psychological trauma experienced by the patients, the development of accurate and rapid diagnostic expert systems is of great importance. Manual inspection of biomedical electroencephalography (EEG) signals and spectrograms is time-consuming, expensive, and often results in eye fatigue. The EEG signals, usually recorded for days, generated a huge amount of data. With the increased size of datasets, more powerful machine learning models, such as deep learning (DL), will greatly improve expert systems and assist physicians in making better-informed decisions. In the literature, a multitude of machine learning models has been proposed to aid neurologists in automating the tasks of epileptic seizure detection and prediction. Despite the growing body of literature on this subject, there is still a gap in knowledge regarding the incorporation of the powerful metaheuristic algorithms in enhancing the DL model. We propose to bridge this gap by leveraging the remarkable success of metaheuristic algorithms in solving optimization and engineering problems. The enhancement in the DL model can be achieved in various phases, such as feature extraction, network architecture, and training algorithms. Publicly available datasets will be pre-processed before being fed as inputs to the improved machine learning models. The performance will be evaluated using statistical performance metrics. This research work will not only improve the lifestyles of epileptic patients and save neurologists' time but also provide a technological innovation that will revitalize the nation's healthcare system. By addressing non-communicable diseases such as epilepsy, the advanced data discovery methods and innovative technology will aid the healthcare industry in producing better-informed decisions, leading to improved patient care.