Areas exhibiting significant changes for self-reward probability and outcome in BOLD signals by GLM analyses
| Activated clusters | Hemi | x | y | z | BA | k | Corrected p |
|---|---|---|---|---|---|---|---|
| Self-reward probability | |||||||
| Control-choice trial | |||||||
| mPFC | LR | 1.3 | 43.2 | 5.9 | 10/32 | 64 | 0.004a |
| Insula | L | −45.6 | 2 | 4.3 | 13/22 | 106 | 0.014 |
| Insula | R | 49.7 | −2.4 | 0.7 | 13/22 | 152 | 0.006 |
| MTG, STG, IPL | L | −52.9 | −45.3 | 18.6 | 22/39/40 | 217 | 0.002 |
| MTG, STG, IPL | R | 56.6 | −34.2 | 21.8 | 22/39/40 | 213 | 0.002 |
| Main trialb | |||||||
| Insula/putamen | L | −49 | 0.1 | 5.1 | 13/22 | 67 | 0.004 |
| Insula | R | 46 | −1.6 | 10.1 | 13 | 19 | 0.033 |
| MTG, STG | L | −54.1 | −43.8 | 12.3 | 22/40 | 40 | 0.016 |
| MTG, IPL | R | 52.9 | −25.8 | 20.4 | 40 | 38 | 0.028 |
| Contrast of control-choice trial > main trials | |||||||
| Insula | R | −49.2 | −0.5 | −7.5 | 13 | 20 | 0.024 |
| MTG, STG | L | −50.5 | −49 | 18 | 40 | 65 | <0.001 |
| MTG, IPL | R | −58.3 | −36.8 | 23.1 | 40 | 85 | <0.001 |
| Contrast of main trials > control-choice trial | |||||||
| No cluster survived at corrected p < 0.05 | |||||||
| Outcome | |||||||
| mPFC | LR | 1.1 | 45.8 | 4.6 | 10/32 | 123 | 0.008 |
| Nucleus accumbens | LR | −0.5 | 10.8 | −2.9 | – | 67 | 0.03 |
| STG | R | 56.9 | −28.2 | 13.4 | 41 | 56 | 0.039 |
-
Activated clusters of variables were observed by voxel-wise GLM analysis of BOLD signals during the decision phase. p < 0.05, FWE corrected by permutation testing with whole-brain voxels (cluster-defining threshold: p < 0.001, otherwise detailed in Notes a and b). Activations by a voxel-wise contrast for main trials versus control-choice trials were obtained [FWE-corrected p < 0.05, for which an uncorrected p = 0.005 was first used and then corrected within the union of activation maps (logical “OR”) between main and control-choice trials]. No cluster survived for the contrast of main trial > control-choice trial for self-reward probability. The stereotaxic coordinates are in Talairach space.
-
↵a Corrected p = 0.004, FWE corrected within activation maps for self-decision values in control-choice trials, uncorrected p = 0.005.
-
↵b The corrected p values reported in this table were based on the FWE correction within the corresponding ROIs of activations in control-choice trials (uncorrected p = 0.01). If corrected within the activation map for self-reward probability in control-choice trials, we have corrected p = 0.016, 0.100, 0.040, and 0.042 for each cluster.
