Fig. 3: Prediction of SANS and SAPS scores.
From: Opposite effects of positive and negative symptoms on resting-state brain networks in schizophrenia

a Four brain network measures, integration strength HIn, segregation strength HSe, integration variability FIn, and segregation variability FSe, were used to predict the SAPS (upper panel) and SANS (lower panel) scores. The bar charts are correlations between real SAPS/SANS scores and predicted scores using different brain measures. ***p < 0.001, **p < 0.01, *p < 0.05. b The correlation between SAPS (upper panel)/SANS (lower panel) scores and predicted scores for segregation variability FSe. c The weights of regions in the best predictions of SAPS (upper panel)/SANS (lower panel) scores were mapped to the brain surface. d The summed weight of the regions in seven functional systems in the best predictions of SAPS (upper panel)/SANS (lower panel) scores (see similar results for the brain parcellation of 500 regions in Supplementary Fig. 8).