Understanding Cognitive Behavioral Therapy for Psychosis Through the Predictive Coding Framework.

Area of Science:

• Cognitive Neuroscience
• Psychiatry
• Computational Neuroscience

Background:

• Psychological treatments like cognitive behavioral therapy for psychosis (CBTp) are effective for persecutory delusions.
• Mechanistic understanding of CBTp often lacks a bridge to cognitive neuroscience.
• Predictive coding offers a framework for understanding psychosis symptoms, including delusions.

Purpose of the Study:

• To link advances in CBTp for persecutory delusions to computational neuroscience.
• To explain CBTp's efficacy using predictive coding and Bayesian learning models.
• To elucidate how CBTp targets information processing deficits in psychosis.

Main Methods:

• Review of recent advances in CBTp for persecutory delusions.
• Application of predictive coding principles, specifically hierarchical Bayesian learning.
• Analysis of prediction error signaling in belief updating.

Main Results:

• Bayesian learning models provide a superior account of CBTp's impact compared to associative models.
• Hierarchical predictive coding explains abnormal belief updating in psychosis via strong priors and noisy data.
• CBTp is shown to systematically alter sensory data selection and interpretation.

Conclusions:

• CBTp strengthens alternative beliefs by targeting information processing at a computational level.
• Predictive coding provides a robust neuroscientific framework for understanding CBTp mechanisms in persecutory delusions.
• Future research should further explore these computational and neuroscientific links.