Back

How the Brain Predicts Timing: Distinct Network Hubs for Predicting and Evaluating Auditory Sensory Events

Nagy, P.; Kovacs, P.; Boncz, A.; Szalardy, O.; Baumgartner, R.; Ignatiadis, K.; Winkler, I.; Toth, B.

2025-11-05 neuroscience
10.1101/2024.11.19.624242 bioRxiv
Show abstract

Temporal prediction enhances perceptual processing by aligning neural excitability with expected sensory events. While local oscillatory mechanisms are known to support timing, less is understood about how large-scale functional brain networks dynamically coordinate predictive processes. In particular, it remains unclear how functional connectivity (FC)--the integration of information into network hubs--differs during expectation formation (post-cue) versus outcome evaluation (post-target), and how this varies across levels of predictability. To investigate this, we recorded electroencephalogram (EEG) while participants performed a cued auditory target-detection task with varying temporal predictability (80% and 50%). Event-related potential (ERP) results revealed that implicit temporal predictability primarily modulated later evaluative processes (P3b, frontal negativity), rather than early sensory components, consistent with context updating under uncertainty. FC was analyzed using a data-driven approach based on Normalized Directed Transfer Entropy (NDTE) applied to EEG difference waveforms between high- and low-predictability conditions. Connectivity was examined separately for the post-cue and post-target periods to distinguish prediction and evaluation phases. Behaviorally, higher temporal predictability facilitated faster reaction times. Connectivity analyses revealed largely overlapping but somewhat distinct network dynamics for prediction and evaluation phases of the signal processing.

Published in Frontiers in Neuroscience (predicted rank #7) · training set

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.