Transfer learning on physics-informed neural networks for tracking the hemodynamics in the evolving false lumen of dissected aorta.

Hyperconnectivity in the cerebello-thalamo-cortical (CTC) circuit, a key component of predictive coding, has been linked to schizophrenia and psychosis risk among youth at clinical high-risk for psychosis (CHR-P). The cerebellum monitors prediction errors, and its dysfunction may lead to misattributions of internal experiences and, consequently, hallucinations and delusions. Neighborhood social fragmentation, characterized by disrupted social ties and unpredictable norms, may compromise the brain's ability to form stable models of the social world. Within predictive coding frameworks, such environments can hinder belief updating, increasing reliance on prior beliefs. This maladaptive process may disrupt CTC connectivity, thereby increasing vulnerability to psychosis. This study examined whether neighborhood social fragmentation predicted CTC connectivity in CHR-P and healthy comparison (HC) youth using functional connectivity data from the North American Prodrome Longitudinal Study Phase 2. Generalized linear mixed models assessed associations between social fragmentation and CTC connectivity, adjusting for age, sex, race/ethnicity, individual poverty, neighborhood educational attainment (proportion of residents with <9th grade education), and study site. Participants (mean [SD] age = 19.46 [4.34]; 42.9 % female; 44.4 % White non-Hispanic) included 115 CHR-P and 74 HCs. CHR-P youth exhibited significantly greater CTC connectivity (mean [SD] = 0.28 [0.91]) compared to HC youth (mean [SD] = -0.44 [0.98]). Greater neighborhood social fragmentation was associated with increased CTC connectivity among CHR-P youth (adjusted β = 0.021, 95 % CI = 0.004-0.038), with a similar, nonsignificant trend among HCs (adjusted β = 0.016, 95 % CI = -0.009-0.041). These findings suggest that the social environment may contribute to psychosis risk by modulating CTC connectivity, highlighting potential targets for prevention and early intervention.