Epileptic Data Classification Using Frequency Power Estimation of Channel (FP1-F7) in Children

Abstract

Epilepsy is the most common neurological disorder in which a person experienced repeated seizures. The seizures are caused by abnormal changes in the electrical and chemical activity in the brain. Early identification and treatment of epilepsy can give good result. About two-third of all children population affected with epilepsy, experience increased seizures with time till their teenage. Conventionally the epileptic data is detected by visual inspection and EEG plot, which greatly depends upon expertise of the examiner. According to latest reports, as high as 23% of total affected population is wrongly diagnosed for epilepsy. Since the antiepileptic medicines approved by FDA, also have numerous side effects, hence it becomes crucial to identify epilepsy correctly. The correct identification is more crucial in where these side effects can cause much damage to the overall development. The authors of this paper have proposed frequency power estimation technique for correct diagnosis of epilepsy using channel(FP1-F7) of EEG data in children below 5 years. The proposed technique correctly detects epileptic cases in children with 90% accuracy.