Sustained EEG responses to rapidly unfolding stochastic sounds reflect Bayesian inferred reliability tracking

Area of Science:

• Neuroscience
• Computational Neuroscience
• Auditory Perception

Background:

• Brain processing of dynamic sensory information relies on integrating sensory input with world expectations.
• Understanding which statistical properties the brain tracks is a key research question.

Purpose of the Study:

• To investigate how the brain processes rapid, stochastic auditory sequences.
• To determine if the brain tracks statistical properties of sound sequences, even when behaviorally irrelevant.

Main Methods:

• Combined electroencephalography (EEG) with computational modeling.
• Participants passively listened to structured tone-pip sequences with varying stochastic patterns.
• A Bayesian predictive inference model guided predictions.

Main Results:

• Listeners automatically tracked sound statistics, irrespective of behavioral relevance.
• Transitions between sound patterns elicited shifts in sustained EEG responses.
• EEG response dynamics correlated with the computational model's tracking of sensory prediction "precision".

Conclusions:

• The brain continuously updates its beliefs about the auditory environment.
• Sustained EEG responses reflect belief updating and the brain's estimation of sensory prediction reliability.
• This study elucidates the interplay between statistical tracking and response dynamics in sensory processing.