Efficient Automatic Techniques for Detection of Neurological Disorders

Electroencephalography (EEG) is used to measure the electric potentials which indicate electrical activity of the subject's brain. The EEG is commonly used by clinicians, researchers, and scientists for studying how the brain works and for identifying brain diseases. The EEG is one of the most vital tools which is used in the study of the patient brain's electrical activity for diagnosing the brain abnormalities like head injury, epilepsy, brain stroke, brain tumors, dementia, sleep disorders, and constant monitoring of the measurement of anesthesia during operation. The work is carried out in this thesis primarily concentrates on how various brain activities of EEG signals can be classified for analyzing several brain neurological diseases. For this, we have proposed methods and algorithms for two-class classification as well as multiclass classification of the EEG signals from different brain activities. The proposed techniques/algorithms are compared with earlier reported work on the same dataset to examine the performance of those algorithms/techniques in terms of classification accuracy.