Disclosure of interest
The authors declare that they have no competing interest.
22 August 2024: Due to technical disruption, we are experiencing some delays to publication. We are working to restore services and apologise for the inconvenience. For further updates please visit our website: https://www.cambridge.org/universitypress/about-us/news-and-blogs/cambridge-university-press-publishing-update-following-technical-disruption
Published online by Cambridge University Press: 15 April 2020
Recently, Jardri and Denève proposed that positive symptoms in schizophrenia could be generated by an imbalance between excitation and inhibition in brain networks, which leads to circular inference, an aberrant form of inference where messages (bottom up and/or top down) are counted more than once and thus, are overweighted [1]. Moreover, they postulated that psychotic symptoms are caused by a system that “expects what it senses” and as a result attributes extreme weight even to weak sensory evidences. Their hypothesis was then validated by a probabilistic inference task (in prep.). Here, we put forward a new experimental study that could validate the circular inference framework in the domain of visual perception. Initially, we restricted ourselves to healthy controls, whose tendencies for psychotic symptoms were measured using appropriate scales. We investigated the computations performed by perceptual systems when facing ambiguous sensory evidence. In those cases, perception is known to oscillate between two interpretations, a phenomenon known as bistable perception. More specifically, we asked how prior expectations and visual cues affect the dynamics of bistability. Participants looked at a Necker cube that was continuously displayed on the screen and reported their percept every time they heard a sound [2]. We manipulated sensory evidence by adding shades to the stimuli and prior expectations by giving different instructions concerning the presence of an implicit bias [3]. We showed that both prior expectations and visual cues significantly affect bistability, using both static and dynamic measures. We also found that the behavior could be well fitted by Bayesian models (“simple” Bayes, hierarchical Bayesian model with Markovian statistics). Preliminary results from patients will also be presented.
The authors declare that they have no competing interest.
Comments
No Comments have been published for this article.