Distinct neuroimaging signatures of OSSO compared to schizophrenia and healthy controls using graph theoretical analysis

Background This study examined topological features and network resilience in schizophrenia spectrum disorders (SSDs), other specified schizophrenia spectrum and other psychotic disorder (OSSO), and healthy controls (HC) with resting-state functional MRI (rs-fMRI) and graph theoretical analysis. Associations between topological metrics, resilience, and symptom severity were also explored. Methods rs-fMRI data from SSDs (n = 77), OSSO (n = 86), and HC (n = 83) were analyzed for global efficiency (Eg), characteristic path length (Lp), nodal local efficiency (NLe), nodal clustering coefficient (NCp), and resilience derived from k-shell decomposition and targeted-attack simulations. Symptom severity was assessed using the Positive and Negative Syndrome Scale (PANSS). Results Both patient groups showed reduced Eg and increased Lp compared with HC, indicating disrupted global integration. At the nodal level, the fusiform gyrus exhibited decreased NLe and NCp in both groups. In OSSO, these nodal metrics correlated with PANSS general and total scores. SSDs displayed pronounced reductions in k-core and maximum-core resilience, whereas OSSO largely retained network stability. k-Shell resilience was most impaired in SSDs, with OSSO showing intermediate deficits. Notably, k-shell resilience in the right superior occipital gyrus significantly differed between OSSO and SSDs. Conclusion This study presents the first investigation of OSSO-specific neuroimaging signatures using network resilience analysis. OSSO showed partial preservation of k-core resilience and intermediate k-shell resilience between SSDs and HC, suggesting distinct neurobiological organization within the psychosis spectrum. k-Shell resilience in the superior occipital gyrus may serve as a potential neuroimaging marker distinguishing OSSO from SSDs.