Prefrontal Cortex Codes Representations of Target Identity and Feature Uncertainty

Area of Science:

• Cognitive Neuroscience
• Neuroscience of Attention

Background:

• Real-world objects have dynamic features, introducing uncertainty crucial for attention-demanding tasks like visual search.
• Understanding how the brain integrates feature uncertainty with target identity for predictive search remains a gap.

Purpose of the Study:

• To investigate how the prefrontal cortex encodes statistical knowledge about search targets, including identity and uncertainty, prior to search onset.
• To determine if target identity and uncertainty are represented independently in the brain.

Main Methods:

• Used functional magnetic resonance imaging (fMRI) to analyze brain activity in 20 human participants during a visual search task.
• Examined multivariate patterns and overall activation in dorsolateral prefrontal cortex (DLPFC) and inferior frontal junction (IFJ) before the search display.
• Analyzed univariate signals in the IFJ in relation to target-stimulus distance and expected variability.

Main Results:

• Observed distinct neural codes for target identity (mean) and uncertainty (variance) in the DLPFC and IFJ before the search display.
• Found that the IFJ signal scaled with target distance from the expected mean, particularly when expected variability was low.
• Demonstrated independent coding of target identity and uncertainty within the same prefrontal regions.

Conclusions:

• The prefrontal cortex represents both the identity and expected variability of features independently.
• These findings advance theories of attention by elucidating how the prefrontal cortex uses identity and uncertainty information for top-down attentional control.