Seizure and Sleeping Disorder Detection for identification of Angelman Syndrome from EEG Using Machine Learning

Abstract

Angelman syndrome (AS) is distinguished by extreme developmental delay or learning disability, gait ataxia, severe speech disorder, and/or tremulousness of the limbs, and distinct activity characterized by repeated laughter, joking, and excitability. Seizures and sleeping disorder are also very common. The earliest signs of developmental delays appear at the age of six months, but the distinctive clinical characteristics of AS do not appear until after one year. A seizure is an electrical disruption in the brain that occurs suddenly and without warning. It may affect the behaviour, growth, and emotions of epileptic patients, as well as their state of consciousness. The ability to anticipate epileptic seizures quickly will save an epileptic patient a lot of trouble, such as slipping, drowning, crashes, and maternity complications. The function extraction stage and the classification stage are the two primary stages of seizure identification. In this article, a new algorithm is introduced for detecting seizures in as little as 10 seconds. To describe the actions of EEG operations, a variety of features are derived from the signal. The classifiers are fed these characteristics. SVM, KNN, and decision tree are the classifiers in question. The findings reveal that SVM is the most accurate and sensitive classifier for predicting the occurrence of seizures.