Common Spatial Pattern Technique With EEG Signals for Diagnosis of Autism and Epilepsy Disorders

Electroencephalogram (EEG) signals reflect the activities or electrical disturbances in neurons human brain. Therefore, these are vital for diagnosing certain brain disorders. This study mainly focused on diagnosis of epilepsy and autism spectrum disorders (ASDs) through analysis processing EEGs. In this study, artifacts were removed from EEG datasets using Independent Component Analysis filtered a fifth-order band-pass Butterworth filter to remove interference noise. Next, new methods used extract features EEGs common spatial pattern (CSP). It is known that conventional CSP uses variance. However, here use entropy, energy, band power with was proposed Then, our investigation, four techniques employed classification, namely, linear discriminant analysis, support vector machine, k-nearest neighbor (KNN), artificial neural network, aim comparing recommending optimal combination ASDs. Finally, effects segment length, frequency band, reduction number results investigated. Two verify methods: King Abdulaziz University dataset (for ASD) MIT epilepsy). The indicated extracted based LBP produced best performance CSP-LBP-KNN provided average classification accuracy approximately 98.46% 98.62% ASDs epilepsy, respectively.