Predictive processing in mental illness: Hierarchical circuitry for perception and trauma.

Area of Science:

• Neuroscience
• Computational Psychiatry
• Cognitive Psychology

Background:

• Predictive coding explains efficient sensory encoding via hierarchical neural circuits passing prediction errors.
• Extensions unify Bayesian inference in neural circuits and psychiatric disorders.
• Perceptual and belief disorders involve weighing prior beliefs against new sensory data.

Purpose of the Study:

• To provide evidence on the development and persistence of psychotic disorders.
• To conceptualize addiction- and trauma-related disorders within the predictive processing framework.
• To reconcile reinforcement learning models with predictive processing and identify application pitfalls for stress.

Main Methods:

• Review and synthesis of recent advances in reinforcement learning momentum models.
• Reconciliation of these models with predictive processing theories.
• Analysis of hierarchical prediction models in the context of stress.

Main Results:

• Trauma-related disorders are increasingly identified as resulting from accumulated prediction errors (uncertainty).
• The issue provides further evidence for predictive processing in psychotic disorders.
• Potential pitfalls in applying hierarchical prediction models to stress are identified.

Conclusions:

• Predictive processing offers a unifying framework for understanding various psychiatric disorders, including trauma-related conditions.
• Future research may require updating purely hierarchical views or reframing long-timescale molecular predictions.
• Reinforcement learning models can be reconciled with predictive processing, offering new insights into stress and related disorders.