Search

Highlights • Schizophrenia classification was studied using a large sample data from eight sites. • Both brain images and genetic factors were used as features for classification. • Machine learning methods and leave-one-site-out cross-validation were performed., Previous brain structural magnetic resonance imaging studies reported that patients with schizophrenia have brain structural abnormalities, which have been used to discriminate schizophrenia patients from normal controls. However, most existing studies identified schizophrenia patients at a single site, and the genetic features closely associated with highly heritable schizophrenia were not considered. In this study, we performed standardized feature extraction on brain structural magnetic resonance images and on genetic data to separate schizophrenia patients from normal controls. A total of 1010 participants, 508 schizophrenia patients and 502 normal controls, were recruited from 8 independent sites across China. Classification experiments were carried out using different machine learning methods and input features. We tested a support vector machine, logistic regression, and an ensemble learning strategy using 3 feature sets of interest: (1) imaging features: gray matter volume, (2) genetic features: polygenic risk scores, and (3) a fusion of imaging features and genetic features. The performance was assessed by leave-one-site-out cross-validation. Finally, some important brain and genetic features were identified. We found that the models with both imaging and genetic features as input performed better than models with either alone. The average accuracy of the classification models with the best performance in the cross-validation was 71.6%. The genetic feature that measured the cumulative risk of the genetic variants most associated with schizophrenia contributed the most to the classification. Our work took the first step toward considering both structural brain alterations and genome-wide genetic factors in a large-scale multisite schizophrenia classification. Our findings may provide insight into the underlying pathophysiology and risk mechanisms of schizophrenia.

A variety of structural abnormalities have been described in post traumatic stress disorder (PTSD), but only a few studies have focused on cortical thickness alterations in recent onset PTSD. In this study, we adopted surface-based morphometry (SBM), which enables an exploration of global structural changes throughout the brain, in order to compare cortical thickness alterations in recent onset PTSD patients, trauma-exposed subjects but without PTSD, and normal controls. Moreover, we used region of interest (ROI) partial correlation analysis to evaluate the correlation among PTSD symptom severity and significant changes of cortical thickness. The widespread cortical thickness reduction relative to the normal controls were found in bilateral inferior and superior parietal lobes, frontal lobes, hippocampus, cingulate cortex, and right lateral occipital lobes in trauma survivors, whereas cortical thickness was only increased in left calcarine cortex in PTSD group. The average cortical thickness of hippocampus and cingulate cortex decreased by 10.75% and 9.09% in PTSD, 3.48% and 2.86% in non PTSD. We further demonstrated that the cortical thicknesses of bilateral ACC and PCC, superior frontal lobes, and hippocampus are negatively correlated with CAPS scores in all trauma survivors. Our study results suggest that stress widens cortical thinning regions and causes more serious effect in recent onset PTSD than non PTSD. It also shows that the cortical thinning in recent onset PTSD predicts the symptom severity., Highlights • PTSD caused by rare, severe disaster (69 miners trapped in 1.4 km underground for 75 h). • The surface-based morphometry is based on such serious, sustained and acute trauma. • The comparisons are among healthy, survivors with, and without recent onset PTSD. • Hippocampus and cingulate thickness in PTSD decreased 3 times than non PTSD. • The altered regions in PTSD group are negatively correlated with CAPS scores.