A Review of machine learning and Deep learning approaches on Schizophrenia Diagnosis

Abstract

Schizophrenia (SCZ) is a serious mental disorder in which people interact with reality abnormally. The symptoms of the diseases are hallucination, delusion, disorganized thinking, etc. Diagnosis of a disease can be done but not accurately. In the long term, it can cause serious consequences and shorten life by more than 10-20 years. Therefore, early and accurate diagnosis of SCZ is common, and methods such as structural magnetic resonance imaging (sMRI), functional MRI (fMRI), diffusion tensor tomography (DTI), and electroencephalogram (EEG) help to demonstrate brain disorders in patients. In addition, during the past decade, researchers have used machine learning (ML) algorithms to obtain an accurate diagnosis of SCZ by distinguishing the brain patterns of healthy and SCZ brains using MRI and fMRI images. Using the presented dataset, machine learning algorithms like Naïve Bayes, Decision Tree, and Random Forest are utilized to forecast the disease. The Python programming language is used for its implementation. Based on accuracy, the researcher presents the optimal algorithm. An algorithm's performance on a given dataset determines its accuracy. This study provides an extensive overview of methods including machine learning classifiers, artificial neural networks (ANN), deep learning (DL) models, methodological foundations, and applications.