• FreeSurfer was used to analyze subcortical shape in
schizophrenia patients.
• Altered surface area and thickness of subcortical
structures were observed.
• The dose of antipsychotic medication did not
significantly influence these changes.
Introduction: Although meta-analysis by combining samples frommultiple studies increases the statistical power, differences in studydesigns, participant characteristics, and outcome measures can createheterogeneity. Therefore, it might be helpful to validate the results ofmeta-analysis in independent smaller single site samples. This studyinvestigates the shape characteristics of subcortical structures inschizophrenia patients by replicating and extending findings from a largemulti-site meta-analysis using a single-site cohort.Methods: Employing the ENIGMA shape analysis protocol, we examinedthe MRI images of 57 patients and 37 controls using FreeSurfer.Results: Compared with controls, schizophrenia was associated withsurface contraction in the left hippocampus, bilateral thalamus, bilateralnucleus accumbens, and right amygdala, alongside surface expansionin the bilateral putamen, left caudate, and bilateral pallidum; the rightcaudate showed both expansion and contraction. Thickness reductionswere observed in the bilateral hippocampus, bilateral thalamus, bilateralaccumbens, and right amygdala, while thickening was most evident inthe right pallidum; the caudate, putamen, and left pallidum showedbidirectional thickness patterns. Antipsychotic dosage showed nosignificant effect on shape measures. Notably, our effect sizes surpassedthose observed in earlier multi-site work. Larger effect sizes observedin our study could be attributed to the higher sampling variability dueto relatively smaller sample size or increased clinical severity amongpatients from our single center compared to the patient populations inthe meta-analysis.Conclusions: Overall, these findings underscore the utility of shape-basedanalyses for understanding schizophrenia neuroanatomy, emphasizeconsidering the clinical characteristics of the study sample, and highlightthe importance of single-site validation—particularly for cohorts notrepresented in larger consortia that have higher baseline clinical severity.Keywords: Antipsychotic, neuroimaging, schizophrenia, shape analysis,subcortical structures