The Dopaminergic Midbrain Encodes the Expected Certainty about Desired Outcomes

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Psychiatry

Background:

• Dopamine's role in learning is established, but its precise function in decision-making and choice remains debated.
• A novel hypothesis posits that dopamine encodes the precision of beliefs about optimal policies, reflecting confidence in chosen actions.

Purpose of the Study:

• To test the hypothesis that dopamine signals the precision of beliefs about optimal policies during decision-making.
• To investigate the neural mechanisms underlying confidence estimation in choice behavior using functional magnetic resonance imaging (fMRI).

Main Methods:

• A novel 'limited offer' game was designed to assess decision-making under uncertainty and risk.
• Functional magnetic resonance imaging (fMRI) was employed to measure midbrain activity, including dopamine projection neurons.
• Bayesian model comparison was used to evaluate behavioral support for active inference versus classical utility maximization.

Main Results:

• Behavior strongly supported active inference (surprise minimization) over utility maximization schemes.
• Midbrain activity, specifically in dopamine projection areas, correlated significantly with model-based estimates of belief precision.
• Trial-by-trial variations in precision estimates accurately predicted neural activity.

Conclusions:

• Human decision-making aligns with hierarchical probabilistic Bayesian inference, involving inference of both optimal policies and their precision.
• Dopamine appears to encode the confidence or precision of these inferences, providing a neural basis for subjective certainty in choices.
• Findings support a unified framework for understanding dopamine function in learning and decision-making through the lens of active inference.