TABLE 3.
Predictions from lesioned models constructed from drug‐naïve patients and validated on drug‐managed patients
| Drug‐naïve group | Drug‐managed group | ||||
|---|---|---|---|---|---|
| r, p permu | Z, p | r, p | Z, p | ||
| Whole brain | .845, .002 | — | .209, .025 | — | |
| Lesioned model | −FP | .811, .001 | 2.149, .032 | .202, .031 | 0.205, .837 |
| −DM | .845, .001 | 0, 1.0 | .202, .030 | 0.535, .592 | |
| −Mot | .809, .002 | 1.426, .154 | .147, .117 | 1.110, .267 | |
| −Vis | .819, .003 | 2.246, .025 | .209, .024 | 0.961, .049 | |
| −Lim | .836, .001 | 0.587, .557 | .195, .036 | 0.639, .523 | |
| −BG | .845, .001 | 0, 1.0 | .192, .039 | 2.232, .026 | |
| −Cer | .843, .001 | 0.695, .392 | .203, .029 | 0.838, .402 | |
Note: Predictability of models constructed from drug‐naïve patients were generated with consensus connections and remained significant by using the optimal threshold of model M (p = .009). Predictability was assessed by calculating Spearman correlation coefficients (r) between observed and predicted scores. The significance of the prediction was confirmed by permutation testing in drug‐naïve patients (p permu). Predictability from whole‐brain matrix is included in the first row for comparison. Results of Steiger's tests did not remain significant after Bonferroni correction.
Abbreviations: BG, basal ganglia network; Cer, cerebellum network; DMN, default mode network; FP, frontoparietal network; Lim, limbic network; Mot, motor network; Vis, visual‐related network.