Analysis Of The Clinical Symptoms Of Schizophrenia Based On FMRI

Schizophrenia(SCZ)is a complex mental disorder,and its incidence rate is about1%.The core symptoms of SCZ are emotional,mental and behavioral disorders.The mortality rate of disability is high,which has caused serious impact and loss on patients and their families.However,at present,the diagnosis and treatment of schizophrenia mainly rely on clinicians’ subjective judgment through the patient’s signs and symptoms,lack of biomarkers for objective evaluation,which is not conducive to prevention,clinical diagnosis and treatment of patients with SCZ.Therefore,it is of great significance to study the pathogenesis of SCZ and find objective biological markers for diagnosis and treatment.With the development of medical imaging,it is possible to achieve the above goals.Functional magnetic resonance imaging(fMRI),which has been widely used in the research of SCZ,has been developing rapidly in recent 20 years and has obtained rich research results.But at present,the research of SCZ mainly focuses on imaging or clinical behavior from a single point of view,and fails to combine the two well.Finding the correlation between the two has become an important research direction.In order to better explore the relationship between imaging and clinical symptoms,this study intends to use sparse canonical correlation analysis(SCCA)to explore the canonical correlation model between the two,and try to find reliable biological markers.This study is divided into three parts: first,using feature selection method,a series of most valuable features are selected from the original features to reduce the original data dimension,so that the performance of subsequent learning algorithm is improved.We use relief(relevant features)to select features.By specifying the number of features to be selected,we select the K features with the largest correlation statistical component.Then,typical correlation patterns between brain biomarkers and clinical diagnostic features are obtained based on SCCA.Here,static functional connectivity is selected for the data associated with biomarkers,which is obtained by brain network modeling of fMRI imaging data.The clinical diagnostic features used data obtained after the psychiatric patients were asked according to the positive and negative syndrome scale.The scale conforms to the principle of psychological measurement,and has standardized evaluation criteria,which can lead to and evaluate the symptoms of mental illness through the exact inquiry scheme.Finally,the typical vector patterns are interpreted from the following aspects.(1)The functional connections and clinical diagnosis features of typical vector pairs were analyzed,and the functional connections were mapped to the corresponding brain nodes at the same time.(2)The relevant functional connections are mapped to the functional brain network system,and the correlation between different functional areas and clinical diagnosis features is analyzed from the perspective of brain functional system.(3)The typical vector pairs are classified by support vector machine(SVM),and the results are analyzed.