Diagnosis of Psychiatric Disorders Using EEG Signal and Machine Learning

This paper explores the use of machine learning in medical diagnosis by an automated diagnosis procedure based on machine learning methodology using electroencephalograph (EEG) data is proposed for diagnosis of psychiatric illness. EEG signal is analyzed in time and frequency domain to extract relevant feature from 32 channel EEG signal. The dimensionality of extracted feature is reduced using feature selection techniques. A variety of classification algorithm is used to classify the class label to achieve high classification. The proposed automated diagnosis system is designed to differentiate various psychiatric conditions, often specific symptoms can appear in more than one diagnostic category, and diagnostic criteria can overlap to the point where confident differentiation is often impossible and adversely affects recovery of the patient. Even the psychiatric expert can have difficulty distinguishing certain psychiatric conditions. The proposed study, demonstrate the use of machine learning methodologies for psychiatric diagnosis with high accuracy to be used in clinical environment.