Gut inference: A computational modelling approach.

Area of Science:

• Neuroscience
• Computational Psychiatry
• Gastroenterology

Background:

• Bayesian inference is theorized to underpin interoception, with prior research focusing on cardiac interoception.
• Recent experimental work has explored heartbeat perception through this lens.
• Extending computational models to other interoceptive domains is a growing area of research.

Purpose of the Study:

• To apply a Bayesian computational model to a novel gastrointestinal interoception task.
• To validate the use of computational modeling for studying interoception beyond the cardiac system.
• To test neural process theories, specifically active inference, within a gastrointestinal context.

Main Methods:

• Utilized a Bayesian computational model applied to a gastrointestinal interoception task.
• Collected simultaneous electroencephalogram (EEG) and peripheral physiological data from 40 healthy participants.
• Analyzed event-related potentials (ERPs) in relation to computational parameters.

Main Results:

• Demonstrated the validity of the Bayesian modeling approach for gastrointestinal interoception.
• Provided confirmatory evidence for active inference theory, linking computational parameters to specific EEG patterns.
• Observed exploratory evidence for computational parameters influencing peripheral physiological regulation.

Conclusions:

• The Bayesian computational approach shows promise for investigating individual differences in gastrointestinal interoception.
• This framework offers a method to link computational parameters to neural activity (EEG) and physiological states.
• The study extends the application of Bayesian inference to the complex domain of gut-brain interactions.