Extracting Reliable Handwriting Kinematic Feauters by using Neural Network for Diagnosing Schizophrenia Disease

Abstract

Schizophrenia (SZ) disease is a kind of severe and rather unknown brain disorder which about one percent of people of the world are affected by the disease. The patients face illusion and severe fear. After emerging the disease's symptoms, the usual way for recognizing the disease and its monitoring during the treatment is a computerized tomography (CT) scan of the brain. The problems like side effects, high cost and leak of high accessibility have caused that finding a new manner instead of the CT scan of brain is considered. Most symptoms found are related to movement (motor symptoms). In this paper, after collecting the data related to kinematic features of pen movements in handwritings of a group affected with Schizophrenia and another group of healthy persons, an artificial neural network (ANN) is used for classification. We discuss how a feed forward ANN can classify data more reliable than Artificial Immune Systems (AIS). Also using ANN the more reliable handwriting kinematic features are extracted for classification. The results show the efficiency of proposed method.