Electrophysiological Based Detection and Classification of Epileptic Seizure using Machine Learning Techniques

Abstract

Epilepsy is a chronic brain disease characterized by repetitive seizures due to erratic electrical discharge causing involuntary behavioral changes such as loss of consciousness and convolutions. Almost 80% of epilepsy patients are in low and middle income countries, where three-fourths of these people face either a treatment gap or a shortage of anti-seizure medicines. As such, the frequency of occurrence of epileptic events are unpredictable, making diagnosis and treatment become difficult. Seizures have four stages: pre-ictal, ictal, post-ictal, and inter ictal. Pre-ictal is just before the occurrence of an epileptic seizure; ictal is the onset period; post-ictal is just after the onset up to 10 minutes; and inter-ictal is after around 10 minutes of onset and lasts till the next occurrence of a seizure. The pre-ictal stage usually involves dizziness, headache, and nausea and is followed by the stage of intense electrical activity in the brain called the ictal region. Then comes the post-ictal region, where the patient returns to baseline conditions along with symptoms like disorientation, drowsiness, and headache. newline