Risk prediction and aversion by anterior cingulate cortex

Area of Science:

• Neuroscience
• Cognitive Psychology
• Computational Psychiatry

Background:

• The error-likelihood hypothesis posits anterior cingulate cortex (ACC) activation correlates with perceived error probability.
• A computational model underpinning this hypothesis has been extended to include the magnitude of potential error consequences.

Purpose of the Study:

• To test the extended computational model's prediction that ACC is sensitive to both error likelihood and consequence magnitude.
• To investigate the neural basis of 'expected risk' in decision-making.
• To explore individual differences in ACC responses related to risk tolerance.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was employed using an incentive-change signal task.
• Analysis focused on ACC activity in relation to predicted error likelihood and consequence magnitude.

Main Results:

• fMRI results replicated the previously observed error-likelihood effect.
• The study validated the computational model's prediction regarding ACC sensitivity to consequence magnitude.
• ACC activation reflected an 'expected risk' calculation (error likelihood × consequence magnitude).
• Error-likelihood and expected risk effects were more pronounced in risk-averse individuals.

Conclusions:

• The findings support an expected risk model of ACC function.
• ACC may contribute to cognitive control by modulating brain activity to mitigate risk.
• Individual differences in risk tolerance influence ACC's sensitivity to error-related signals.