Learning in Visual Regions as Support for the Bias in Future Value-Driven Choice

Area of Science:

• Neuroscience
• Cognitive Science
• Decision Making

Background:

• Reinforcement learning biases decisions towards high-value options.
• This bias is linked to stimulus value tracking and outcome evaluation in the striatum and prefrontal cortex.
• The interplay between learning and sensory perception in decision-making is not fully understood.

Purpose of the Study:

• To investigate the relationship between visual brain activity and value-based decision-making.
• To explore how visual processing supports learning and choice biases.
• To determine if visual cortex activity predicts future value-driven choices.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was used to measure blood oxygen level-dependent (BOLD) responses in 43 participants.
• BOLD responses in visual, motor, prefrontal, and striatal regions were analyzed in relation to value beliefs and prediction errors.
• A supervised machine learning algorithm was employed to decode future choice outcomes based on trial-by-trial BOLD data.

Main Results:

• Visual blood oxygen level-dependent (BOLD) responses correlated with value beliefs during choice and prediction errors after outcomes.
• Machine learning successfully decoded future choice outcomes with 70% accuracy using BOLD data from visual and other brain regions.
• Visual activity was a significant predictor of future value-driven choices, alongside activity in striatal and prefrontal regions.

Conclusions:

• Value tracking occurs not only in the striatum and prefrontal cortex but also in the visual cortex.
• Visual cortex activity plays a crucial role in supporting the brain's bias towards more valued options.
• These findings suggest a more integrated role for sensory processing in value-based decision-making than previously thought.