Regional amplitude of low-frequency fluctuation (ALFF) abnormalities in schizophrenia are often interpreted one region at a time, even though the implicated areas differ in effect magnitude, spatial extent, anatomical position, and classification behavior. The present study asked whether these dimensions converge on a stable distributed phenotype and whether that phenotype is concentrated in posterior association cortex. Resting-state functional magnetic resonance imaging measurements from 131 adults with schizophrenia and 128 healthy participants yielded seven regions with lower ALFF: bilateral precuneus, left angular gyrus, bilateral middle temporal gyri, left anterior cingulate cortex, left fusiform gyrus, and right cerebellum. Acquisition, preprocessing, voxelwise comparison, support vector classification, and receiver-operating-characteristic procedures were retained. A topology-constrained regional evidence method was added. It combined percentile scores for attenuation magnitude, cluster extent, peak classification accuracy, and Gaussian distance-kernel centrality. Equal-weight scores were examined together with 200,000 Dirichlet-distributed weight vectors, leave-one-criterion analyses, posterior evidence-share estimation, summary-statistic demographic tests, and decision-balance calculations. The bilateral precuneus ranked first (score 0.875) and the left angular gyrus ranked second (0.768). The precuneus occupied rank 1 in 79.4% of weight draws, while the angular gyrus did so in 20.0%. Posterior association regions accounted for 70.5% of equal-weight evidence; their median share over the weight simplex was 70.8% (central 95% range 64.5–75.8%). Excluding any one criterion preserved the ordering strongly (Spearman ρ = 0.883–0.982). The patient group was younger by 3.99 years (Hedges g = −0.43), whereas education was comparable. Precuneus and angular-gyrus decision profiles had nearly identical balanced accuracy (73.15% and 73.12%) despite different sensitivity–specificity combinations. Schizophrenia-related ALFF attenuation is best characterized as a stable posterior associative pattern led by the precuneus and angular gyrus, not as an isolated single-locus effect. Its regional ordering is robust to evidence preferences, but demographic imbalance and modest specificity restrict individual diagnostic interpretation.
Regional amplitude of low-frequency fluctuation (ALFF) abnormalities in schizophrenia are often interpreted one region at a time, even though the implicated areas differ in effect magnitude, spatial extent, anatomical position, and classification behavior. The present study asked whether these dimensions converge on a stable distributed phenotype and whether that phenotype is concentrated in posterior association cortex. Resting-state functional magnetic resonance imaging measurements from 131 adults with schizophrenia and 128 healthy participants yielded seven regions with lower ALFF: bilateral precuneus, left angular gyrus, bilateral middle temporal gyri, left anterior cingulate cortex, left fusiform gyrus, and right cerebellum. Acquisition, preprocessing, voxelwise comparison, support vector classification, and receiver-operating-characteristic procedures were retained. A topology-constrained regional evidence method was added. It combined percentile scores for attenuation magnitude, cluster extent, peak classification accuracy, and Gaussian distance-kernel centrality. Equal-weight scores were examined together with 200,000 Dirichlet-distributed weight vectors, leave-one-criterion analyses, posterior evidence-share estimation, summary-statistic demographic tests, and decision-balance calculations. The bilateral precuneus ranked first (score 0.875) and the left angular gyrus ranked second (0.768). The precuneus occupied rank 1 in 79.4% of weight draws, while the angular gyrus did so in 20.0%. Posterior association regions accounted for 70.5% of equal-weight evidence; their median share over the weight simplex was 70.8% (central 95% range 64.5–75.8%). Excluding any one criterion preserved the ordering strongly (Spearman ρ = 0.883–0.982). The patient group was younger by 3.99 years (Hedges g = −0.43), whereas education was comparable. Precuneus and angular-gyrus decision profiles had nearly identical balanced accuracy (73.15% and 73.12%) despite different sensitivity–specificity combinations. Schizophrenia-related ALFF attenuation is best characterized as a stable posterior associative pattern led by the precuneus and angular gyrus, not as an isolated single-locus effect. Its regional ordering is robust to evidence preferences, but demographic imbalance and modest specificity restrict individual diagnostic interpretation.
