29.1 COMPUTATIONAL MODELING OF PERCEPTION AND BEHAVIOR REVEALS HIDDEN INSIGHTS INTO MECHANISMS OF PSYCHOTIC SYMPTOMS. (9th April 2019)

• Record Type: Journal Article
• Authors: Powers, Albert; Mathys, Christoph; Corlett, Philip
• Is Part Of: Schizophrenia bulletin. Volume 45(2019)Supplement 2
• Page Start: S136
• Page End: S136
• Publication Date: 2019-04-09
• Subjects: Schizophrenia -- Periodicals; Schizophrenia -- Research -- Periodicals; 616.898005
• DOI: 10.1093/schbul/sbz022.117
• Languages: English
• ISSNs: 0586-7614

Abstract

Abstract: Background: Because it happens so quickly and automatically, it may be intuitive to think of perception as a passive process in which our brains simply receive input about our surroundings via our sensory organs. However, it has long been clear that perception is, instead, an active process, characterized by the building of an internal model of our environment, blending incoming sensory evidence with prior beliefs about what is around us. Within this framework, hallucinations may be thought to arise from an increased influence of these prior beliefs during perception. Methods: To test this idea, we adapted a classical sensory conditioning paradigm to the functional imaging setting: participants were exposed to repeated pairings of a salient visual stimulus with a faintly-presented auditory stimulus and subsequently reported the perception of the auditory stimulus even when it was not present, contingent on the presence of the visual—a conditioned hallucination. We recruited participants who differed orthogonally in whether they had hallucinations and/or psychosis, resulting in four groups: those with psychosis both with hallucinations (P+H+) and without (P+H-), otherwise healthy voice-hearers (P-H+), and healthy controls (P-H-). Results: We found that conditioned hallucinations readily occur in all subjects but with markedly increased frequency in those who hallucinate (P+H+ and P-H+), regardless of psychosis status. Conditioned hallucinations activated stimulus-responsive auditory cortex in addition to a network of regions active during clinical hallucinations. Computational modeling demonstrated an increased reliance on heightened prior beliefs in those with hallucinations (encoded by anterior insula and superior temporal sulcus), regardless of psychosis status. By contrast, those with psychosis exhibited a decreased recognition of changing stimulus contingencies (encoded by cerebellum and hippocampus), regardless of hallucination status. Conclusions: These results may represent an objective means to distinguish people with hallucinations from those without, and, orthogonally, a need for treatment from those without. Preliminary data indicate these measures may also be useful in risk stratification in those at clinical high risk (CHR) of psychosis. Future work will focus on characterizing the predictive power of these measures in both diagnosis and treatment selection.