Neural coding of reward-prediction error signals during classical conditioning with attractive faces

Area of Science:

• Neuroscience
• Psychology
• Behavioral Economics

Background:

• Attractive faces are recognized as visual rewards, activating brain regions like the orbitofrontal cortex and nucleus accumbens.
• Previous research indicates reward structures are involved in processing attractive faces.

Purpose of the Study:

• To investigate if arbitrary neutral stimuli can acquire value through conditioning with attractive faces.
• To examine the neural basis of this conditioning using functional magnetic resonance imaging (fMRI).
• To determine if a reward-prediction error signal is involved in learning with attractive faces.

Main Methods:

• Human subjects underwent a conditioning procedure pairing neutral stimuli with attractive or unattractive faces.
• fMRI was used to scan brain activity during the conditioning process.
• A Rescorla-Wagner learning model generated a reward-prediction error signal for regression analysis with fMRI data.

Main Results:

• Behavioral ratings showed changes in conditioned stimuli (CS) value after pairing, particularly with attractive female faces.
• Significant reward-prediction error-related activity was observed in the ventral striatum during conditioning with attractive faces compared to unattractive ones.
• fMRI data revealed prediction error signals in the ventral striatum, a key dopamine pathway target.

Conclusions:

• Arbitrary stimuli can gain conditioned value by association with pleasant visual stimuli, similar to other rewards like money or juice.
• This learning process engages a reward-prediction error signal in the ventral striatum.
• Findings may inform advertising strategies that use associations between products and attractive visuals to influence consumer preferences.