Human brain integrates both unconditional and conditional timing statistics to guide expectation and behavior.

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience

Background:

• The brain predicts upcoming event timings using hazard functions based on probability distributions.
• Integrating multiple temporal probability distributions for coherent prediction remains unclear.

Purpose of the Study:

• Investigate how the brain integrates unconditional and conditional temporal predictions.
• Identify neural mechanisms underlying hierarchical temporal prediction.

Main Methods:

• Developed a foreperiod sequence paradigm with paired trials.
• Modeled reaction times using hazard functions representing unconditional and conditional predictions.
• Analyzed electroencephalographic (EEG) source signals.

Main Results:

• Behavioral models integrating both prediction types best explained reaction times.
• EEG source signals were best reconstructed by integrating both predictions.
• Unconditional and conditional predictions are encoded in posterior and anterior regions, respectively.
• The right posterior cingulate area integrates these predictions.

Conclusions:

• The brain integrates multilevel temporal information for prediction.
• Distinct neural networks support hierarchical predictive coding of time.
• This provides insight into the brain's temporal processing capabilities.