Schizophrenia classification using machine learning on resting state EEG signal

Schizophrenia is a severe mental disorder associated with wide spectrum of cognitive and neurophysiological dysfunctions. Early diagnosis still difficult based on the manifestation disorder. In this study, we have evaluated whether machine learning techniques can help in schizophrenia, proposed processing pipeline order to obtain classifiers schizophrenia resting state EEG data. We computed well-known linear non-linear measures sliding windows data, selected those which better differentiate between patients healthy controls, combined them through principal component analysis. These components were finally used as features five standard algorithms: k-nearest neighbours (kNN), logistic regression (LR), decision trees (DT), random forest (RF) support vector machines (SVM). Complexity showed high level ability differentiating from controls. differences groups mainly located delimited zone right brain hemisphere, corresponding opercular area temporal pole. Based under curve parameter receiver operating characteristic analysis, obtained classification power almost all algorithms tested: SVM (0.89), RF (0.87), LR (0.86), kNN (0.86) DT (0.68). Our results suggest that data able easily compute select set allow perform very efficiently subjects.