The Xi'an Schizophrenia Imaging Lab (SIL) data and ten years of MRI study on schizophrenia

Mounting evidence suggests that function and connectivity of the striatum is disrupted in schizophrenia1-5. We have developed a new hypothesis-driven neuroimaging biomarker for schizophrenia identification, prognosis and subtyping based on functional striatal abnormalities (FSA). FSA scores provide a personalized index of striatal dysfunction, ranging from normal to highly pathological. Using inter-site cross-validation on functional magnetic resonance images acquired from seven independent scanners (n = 1,100), FSA distinguished individuals with schizophrenia from healthy controls with an accuracy exceeding 80% (sensitivity, 79.3%; specificity, 81.5%). In two longitudinal cohorts, inter-individual variation in baseline FSA scores was significantly associated with antipsychotic treatment response. FSA revealed a spectrum of severity in striatal dysfunction across neuropsychiatric disorders, where dysfunction was most severe in schizophrenia, milder in bipolar disorder, and indistinguishable from healthy individuals in depression, obsessive-compulsive disorder and attention-deficit hyperactivity disorder. Loci of striatal hyperactivity recapitulated the spatial distribution of dopaminergic function and the expression profiles of polygenic risk for schizophrenia. In conclusion, we have developed a new biomarker to index striatal dysfunction and established its utility in predicting antipsychotic treatment response, clinical stratification and elucidating striatal dysfunction in neuropsychiatric disorders.

Past work on relatively small, single-site studies using regional volumetry, and more recently machine learning methods, has shown that widespread structural brain abnormalities are prominent in schizophrenia. However, to be clinically useful, structural imaging biomarkers must integrate high-dimensional data and provide reproducible results across clinical populations and on an individual person basis. Using advanced multi-variate analysis tools and pooled data from case-control imaging studies conducted at 5 sites (941 adult participants, including 440 patients with schizophrenia), a neuroanatomical signature of patients with schizophrenia was found, and its robustness and reproducibility across sites, populations, and scanners, was established for single-patient classification. Analyses were conducted at multiple scales, including regional volumes, voxelwise measures, and complex distributed patterns. Single-subject classification was tested for single-site, pooled-site, and leave-site-out generalizability. Regional and voxelwise analyses revealed a pattern of widespread reduced regional gray matter volume, particularly in the medial prefrontal, temporolimbic and peri-Sylvian cortex, along with ventricular and pallidum enlargement. Multivariate classification using pooled data achieved a cross-validated prediction accuracy of 76% (AUC = 0.84). Critically, the leave-site-out validation of the detected schizophrenia signature showed accuracy/AUC range of 72-77%/0.73-0.91, suggesting a robust generalizability across sites and patient cohorts. Finally, individualized patient classifications displayed significant correlations with clinical measures of negative, but not positive, symptoms. Taken together, these results emphasize the potential for structural neuroimaging data to provide a robust and reproducible imaging signature of schizophrenia. A web-accessible portal is offered to allow the community to obtain individualized classifications of magnetic resonance imaging scans using the methods described herein.

Purpose To investigate auditory verbal hallucination (AVH)-specific patterns of brain activity within the resting-state networks (RSNs) that have been proposed to underpin the neural mechanisms of schizophrenia (SZ). Materials and Methods This cross-sectional study was approved by the local ethics committee, and written informed consent was obtained from all participants prospectively recruited. Independent component analysis was used to investigate RSNs in 17 patients with first-episode untreated SZ with AVHs, 15 patients with SZ without AVHs, and 19 healthy control subjects who underwent resting-state functional magnetic resonance imaging. Dual regression was implemented to perform between-group analysis. Regional brain function was then explored within RSNs by using the amplitude of low-frequency fluctuation. Two-sample t tests were used to compare regional brain function between the two patient groups, and Pearson correlation analysis was used to characterize the relationship between imaging findings and severity of AVHs. Receiver operating characteristic analysis was used to evaluate the diagnostic performance of these brain function measures. Results Independent component analysis demonstrated symptom-specific abnormal disrupted coactivation within the auditory, default mode, executive, motor, and frontoparietal networks and was pronounced in the auditory cortex, supramarginal gyrus, insula, putamen, dorsolateral prefrontal cortex, angular gyrus, precuneus, and thalamus (P < .05 with false discovery rate correction). Amplitude of low-frequency fluctuation analysis demonstrated similar patterns within these RSNs (P < .05 with false discovery rate correction). Furthermore, a positive correlation between the degree of coactivation within the motor network and the severity of AVHs was observed in patients with SZ with AVHs (r = 0.67, P = .003). The area under the receiver operating characteristic curve was 0.76-0.90 for all RSNs. Conclusion These findings indicate that dysfunctional brain regions are involved in auditory processing, language production and monitoring, and sensory information filtering in patients with SZ with AVHs, which may be helpful in furthering the understanding of pathophysiological correlates of AVHs in SZ. Online supplemental material is available for this article.