This prognostic study evaluates whether psychosis transition can be predicted in patients with clinical high-risk syndromes or recent-onset depression by multimodal machine learning that optimally integrates clinical and neurocognitive data, structural magnetic resonance imaging, and polygenic risk scores for schizophrenia.

published version features the final layout of the paper including the volume, issue and page numbers. Link to publication General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the "Taverne" license above, please follow below link for the End User

No abstract

Background Early psychosis in first-episode psychosis (FEP) and clinical high-risk (CHR) individuals has been associated with alterations in mean regional measures of brain morphology. Examination of variability in brain morphology could assist in quantifying the degree of brain structural heterogeneity in clinical relative to healthy control (HC) samples. Methods Structural magnetic resonance imaging data were obtained from CHR (n = 71), FEP (n = 72), and HC individuals (n = 55). Regional brain variability in cortical thickness (CT), surface area (SA), and subcortical volume (SV) was assessed with the coefficient of variation (CV). Furthermore, the person-based similarity index (PBSI) was employed to quantify the similarity of CT, SA, and SV profile of each individual to others within the same diagnostic group. Normative modeling of the PBSI-CT, PBSI-SA, and PBSI-SV was used to identify CHR and FEP individuals whose scores deviated markedly from those of the healthy individuals. Results There was no effect of diagnosis on the CV for any regional measure (P > .38). CHR and FEP individuals differed significantly from the HC group in terms of PBSI-CT (P < .0001), PBSI-SA (P < .0001), and PBSI-SV (P = .01). In the clinical groups, normative modeling identified 32 (22%) individuals with deviant PBSI-CT, 12 (8.4%) with deviant PBSI-SA, and 21 (15%) with deviant PBSI-SV; differences of small effect size indicated that individuals with deviant PBSI scores had lower IQ and higher psychopathology. Conclusions Examination of brain structural variability in early psychosis indicated heterogeneity at the level of individual profiles and encourages further large-scale examination to identify individuals that deviate markedly from normative reference data.

No abstract