All that glitters ... dissociating attention and outcome expectancy from prediction errors signals

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Computational Neuroscience

Background:

• Neural correlates of prediction errors (PEs) were initially identified in dopamine neurons.
• Subsequent research indicated PEs in other brain regions, including the orbitofrontal cortex (OFC), ventral striatum (VS), and amygdala.
• Interpreting neural activity in these regions requires distinguishing PE signals from other cognitive processes.

Purpose of the Study:

• To review the literature on neural activity in the OFC, VS, and amygdala in relation to prediction errors.
• To differentiate between prediction error signaling and other functions like attention and outcome expectancy in these brain areas.
• To clarify the roles of these regions in learning and behavior.

Main Methods:

• Literature review of studies investigating neural activity in dopamine neurons, OFC, VS, and amygdala.
• Analysis of findings related to prediction errors, attention, and outcome expectancy.
• Comparative analysis of neural activity patterns across different brain regions.

Main Results:

• Dopamine neurons show activity consistent with prediction error signaling.
• Basolateral amygdala (BLA) activity to unexpected outcomes likely reflects attentional processes.
• Orbitofrontal cortex (OFC) and ventral striatum (VS) activity is not modulated by prior expectations and may signal outcome expectancy.

Conclusions:

• Neural activity in the OFC, VS, and amygdala does not uniformly represent prediction errors.
• The BLA's role in processing unexpected outcomes may be primarily attentional.
• OFC and VS activity provides information about outcome expectancy, distinct from prediction error signaling.
• These findings refine our understanding of how different brain regions interact to guide learning and behavior.