Survey on Effective Deep Learning-based Automated Epileptic Seizure Detection

Uncontrolled neuron activity in the brain is the source of the neurological disorder known as epilepsy seizures that substantially reduce patients' quality of life and frequently risk their lives. Research on automatic seizure identification is essential for the field of seizure treatment and is necessary for seizure prevention. For these applications to be successful, the classification and detection algorithms must be accurate. Numerous methods have been studied, developed, and introduced over time. This review describes a variety of seizure detection techniques from the last six years, including both established methods and the latest deep learning techniques. Performance comparisons between the advantages and disadvantages of each method under study were made. This survey revealed that the recent research has mainly examined the strength of Deep Learning methods in identifying and categorizing seizures, which have been employed in a variety of fields such as image processing, signal processing, etc. Deep learning hybrid methods has also been the subject of studies. Finally, the research study is concluded by discussing potential future research directions.