Neural Mechanisms of the Transformation from Objective Value to Subjective Utility: Converting from Count to Worth

• Published in: Frontiers in neuroscience 2016-11, Vol.10, p.507-507
• Main Authors: Kurnianingsih, Yoanna A, Mullette-Gillman, O'Dhaniel A
• Format: Article
• Language: English
• Subjects: Brain; Computational neuroscience; Deactivation; Decision Making; Expected values; fMRI; Frontal gyrus; functional connectivity; Neural networks; Neuroscience; Nucleus accumbens; Preferences; Replication; risk; risky decision making; uncertainty
• ISSN: 1662-4548; 1662-453X; 1662-453X
• DOI: 10.3389/fnins.2016.00507

When deciding, we aim to choose the "best" possible outcome. This is not just selection of the option that is the most numerous or physically largest, as options are translated from objective value (count) to subjective value (worth or utility). We localized the neural instantiation of the value-to-utility transformation to the dorsal anterior midcingulate cortex (daMCC), with independent replication. The daMCC encodes the context-specific information necessary to convert from count to worth. This encoding is not simply a representation of utility or preference, but the interaction of the two. Specifically, the relationship of brain activation to value is dependent on individual preference, with both positive and negative slopes across the population depending on whether each individual's preference results in enhancement or diminishment of the valuation. For a given value, across participants, enhanced daMCC activation corresponds to diminished subjective valuation, deactivation to enhanced subjective valuation, and non-modulated activation with non-modulated subjective valuation. Further, functional connectivity analyses identified brain regions (positive connectivity with the inferior frontal gyrus and negative connectivity with the nucleus accumbens) through which contextual information may be integrated into the daMCC and allow for outputs to modulate valuation signals. All analyses were replicated through an independent within-study replication, with initial testing in the gains domain and replication in the intermixed and mirrored losses trials. We also present and discuss an ancillary finding: we were unable to identify parametric value signals for losses through whole-brain analyses, and ROI analyses of the vmPFC presented non-modulation across loss value levels. These results identify the neural locus of the value-to-utility transformation, and provide a specific computational function for the daMCC in the production of subjective valuation through the integration of value, context, and preferences.