How the brain predicts timing: distinct network hubs for predicting and evaluating auditory sensory events

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Temporal prediction is crucial for efficient perceptual processing by aligning neural activity with expected events.
• Understanding how large-scale brain networks dynamically coordinate predictive processes, particularly functional connectivity (FC) during expectation formation versus outcome evaluation, remains an area of active research.
• The influence of varying predictability levels on these network dynamics is not fully elucidated.

Purpose of the Study:

• To investigate how functional connectivity (FC) differs between the prediction phase (expectation formation) and the evaluation phase (post-target outcome assessment).
• To examine how these network dynamics vary across different levels of temporal predictability in an auditory task.
• To elucidate the role of large-scale brain networks in temporal prediction and outcome evaluation.

Main Methods:

• Electroencephalography (EEG) data were recorded from participants performing a cued auditory target-detection task.
• Temporal predictability was manipulated (80% vs. 50%).
• Functional connectivity (FC) was analyzed using Normalized Directed Transfer Entropy (NDTE) on EEG difference waveforms, distinguishing prediction and evaluation periods.

Main Results:

• Behaviorally, increased temporal predictability led to faster reaction times.
• Event-related potential (ERP) analyses indicated that temporal predictability modulated later evaluative processes (P3b, frontal negativity), not early sensory components.
• Functional connectivity analyses showed a shift from a fronto-temporo-parietal network during prediction to a more focal auditory-frontal circuit during outcome evaluation.

Conclusions:

• Temporal prediction and evaluation involve dynamic interactions within multiple large-scale brain networks.
• These findings support models of predictive processing that involve both distributed and frontal-dominant integration mechanisms.
• The study highlights the dynamic nature of brain network engagement during predictive processing under varying levels of certainty.