Neurological disorder detection using EEG signal processing and Machine Learning

Abstract

Neurological disorders are abnormal behavior of nervous system occurring due to irregular firing of neurons. These disorders cause both physical and psychological imbalance to human being suffering from it and may cause even death in some cases. Few of these disorders are epilepsy, Alzheimer’s disease, dementia, cerebro vascular diseases including stroke, migraine, Parkinson’s disease and many more. This manuscript presents neurological disorder detection using electroencephalogram (EEG) signals with machine learning methods. Here, neurological disorders like epilepsy and Attention deficit/hyperactivity disorder (ADHD) have discussed. These Neurological disorders can be differentiated from normal healthy brain using EEG Signal features and efficient classification methods ANN, SVM, Random Forrest and Ensemble methods etc. Some of the robust features like, RMS value, entropy and wavelet coefficients have been explored. For seizures, epilepsy, and ADHD patients time frequency features like wavelet coefficients are the robust one. One of the databases utilized in this study for epilepsy detection is BONN dataset.