Temporal prediction errors modulate cingulate-insular coupling.

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Prediction error is fundamental for adaptive behavior and learning.
• The anterior mid-cingulate cortex (aMCC) and anterior insula (aINS) are implicated in processing prediction errors, especially under uncertainty.
• Understanding the neural circuitry of temporal prediction is essential for cognitive neuroscience.

Purpose of the Study:

• To investigate the effective connectivity between the aMCC, aINS, and striatum during temporal prediction tasks.
• To model how task context and temporal prediction errors influence neural coupling within this system.
• To determine the specific role of prediction errors versus other factors like attention or task difficulty in modulating aMCC-aINS activity.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was used to scan 15 participants performing a temporal prediction task.
• Participants predicted the timing of a visual event involving moving balls, with uncertainty induced by spatial gaps and temporal delays.
• Dynamic causal modeling (DCM) and Bayesian fMRI analyses were employed to assess effective connectivity.

Main Results:

• The aMCC-aINS system, along with the striatum, is active during both the prediction and occurrence of dynamic events.
• The insula serves as the initial entry point in a three-region pathway for processing temporal predictions.
• Prediction errors significantly influence the aMCC-aINS system, weakening the effect of the aMCC on the aINS, independent of attentional demands, task difficulty, or duration.

Conclusions:

• Temporal prediction errors, rather than task-related factors, are the primary drivers of activity modulation in the aMCC-aINS pathway.
• The study provides a computational model for understanding the neural basis of temporal prediction errors in dynamic tasks.
• This research elucidates the insula's role as a critical hub in processing temporal predictions and prediction errors.