Certain investigations on epileptic seizure detection and classification using hybrid metaheuristic algorithms in artificial neural networks

Abstract

Epilepsy is a chronic illness that causes unpredictable, jarring disruptions in newlinethe brain activity that interfere with an epileptic patient s regular everyday activities. It newlineis one of the most widespread neurological conditions, affecting 65 million people newlineworldwide, ranging in age from infants to the elderly. In emerging countries like India, newlinethis count is getting worse. According to a report from the Indian epilepsy Centre, newlinebetween 0.5 and 1 million new patients are diagnosed each year, resulting in an newlineestimated 15 million Indians living with the disorder. Therefore, precisely identifying newlineand diagnosing epileptic seizures is extremely vital in order to provide patients with newlinemore effective prevention and treatment. This fact has emphasized the need of newlineautomated techniques for diagnosing epilepsy at early stage. Extensive progress has newlinebeen made in these directions, it is still hard to diagnose epilepsy in its earliest stages newlinewith efficiency and accuracy. newlineScreening is an effective way to detect and diagnose epilepsy. newlineElectroencephalogram (EEG) is the most popular and widely used imaging modalities newlinefor epileptic seizure detection and diagnosis. Physicians or Neurologists typically newlineanalyze EEG signals to identify the pattern changes brought on by epileptic seizure newlinesignals. Visual scanning of EEG signal is a very tough and time consuming task that newlinefrequently results in disagreements between analysts. A fully automated system must newlinebe created in order to reduce the need for manual interpretation and to perform faster, newlinemore accurate signal analysis newline