A Computational Psychiatry Approach to Understanding the Brain Mechanism Underpinning Altered Perception in Psychosis

Overview:

Researchers will use brain scans and decision-making tasks to determine if people with psychosis have a different balance of prior beliefs and sensory information when forming perceptions, which could lead to new treatments.

Abstract:

People integrate two sources of information when forming perceptions: sensory information and prior knowledge. As a result, a perceptual experience can be biased by one’s prior beliefs or expectations. Recent work provides contradictory evidence for whether psychotic symptoms arise from an overreliance on prior or on sensory information. We will test these competing hypotheses in the psychosis spectrum in the general and clinical populations and ask whether such alterations occur in a continuous or discontinuous manner within that spectrum. Participants will complete a series of perceptual decision-making tasks under uncertainty while their brain activity is recorded with optically pumped magnetometers (OP-MEG) and while in the 7T MRI. Tasks will be completed by four groups: people on the schizophrenia spectrum (SZS), first-episode psychosis (FEP), ultra-high risk (UHR), and neurotypical individuals (NT) in the general community. Unlike previous work, we will quantify the degree of reliance in both prior and sensory information using Bayesian modelling. Ultra-high field fMRI at 7T has cortical laminar resolution and will elucidate whether priors and likelihoods are encoded in top-down (deep layers) or bottom-up connections (middle layers). Brain connectivity techniques (e.g., Dynamic Causal Modelling) with OP-MEG will uncover the neural dynamics underpinning aberrant perception with unprecedented spatio-temporal resolution. Using a multi-pronged approach (across sensory modalities and brain imaging approaches) we will determine the neural mechanism underpinning altered perception in psychosis. Importantly, knowing whether perceptual alterations in psychosis are caused by aberrant beliefs (top-down processes) or sensations (bottom-up), or both, will inform when therapeutic strategies should focus on cognitive training (to change beliefs), drug interventions (to change sensory encoding), or both. Identifying the fundamental neural mechanisms that cause psychosis will inform rehabilitation strategies by pointing to brain stimulation and pharmacological targets. Moreover, understanding the neural trajectory of psychosis across illness stages will provide a means to predict and prevent transition to severer stages.

Broader Impact:

This project will use cutting-edge brain imaging and decision-making tasks to uncover how people with psychosis process sensory information and beliefs differently, potentially paving the way for new treatments that are tailored to the brain’s underlying mechanisms. By identifying whether altered perception in psychosis stems from disrupted sensory input, mistaken beliefs, or both, the research could help guide more effective interventions and early prevention strategies.