Fig. 4: A network of brain regions tracks the prospective value difference during choices.
From: Grid-like entorhinal representation of an abstract value space during prospective decision making
a Modulation of activity by the difference between model-derived chosen vs. unchosen value during choices. Clusters depicted survive whole-brain correction (two-sided non-parametric permutation test with TFCE and pFWE < 0.05). Statistical image is displayed on the MNI template. b Based on our expectation of a value difference effect in vmPFC, we visualize the effect in the vmPFC cluster by showing the time courses of the effect sizes of the chosen and unchosen value, time-locked to choice onset (choice onset at 0 s). Lines represent the mean across participants, with shaded regions as the 95% confidence interval. c The value difference effect in vmPFC correlates significantly positively with performance (two-sided Pearson correlation, p = 0.02). Dots represent data from n = 46 participants; line represents linear regression line, with shaded regions as the 95% confidence interval. d Modulation of activity by the prospective component of the value difference during choices. The prospective component refers to the influence of values estimated by the prospective Rescorla–Wagner model over values estimated by the original (non-prospective) Rescorla–Wagner model. Clusters depicted survive whole-brain correction (two-sided non-parametric permutation test with TFCE and pFWE < 0.05). Statistical image is displayed on the MNI template. Source data are provided as a Source Data file.
