Fig. 2: The longitudinal changes in gyrification and its negative correlation with treatment response.

A The regions and percentage of longitudinal changes in LGI of patients; The method used in the above results is a one-sample t-test of [follow up - baseline]/baseline against null and paired t-test; in violin plot, each data point refers to the actual percentage of longitudinal change averaged across the cluster representing the labeled region from the vertex-wise analysis (i.e., a spatially contiguous cluster). The distribution across either side of null reflects the observation that some subjects had an increase while others had a decrease over time on average, with the effect sizes in Table 2 indicating a relatively minor magnitude of time-varying changes despite their statistical significance in the within-subject analysis; B Cortical statistical maps displaying the negative correlation between the longitudinal gyrification changes and treatment response. Clusters in cold blue indicate that a longitudinal reduction in LGI in these regions relates to less symptomatic improvement in total, positive and negative symptom scores in the sub-panels of cortical surfaces displayed from top to bottom. The horizontal bar shows p values after correction for multiple comparisons via Monte-Carlo simulations (threshold = 0.01), where the blue region shows Sch < HC with p values from 0.05 to 0.00001 and the red region shows Sch > HC with p values from 0.05 to 0.00001.