Abstract
Psychosis involves abnormal sensory percepts, including hallucinations and more subtle differences in visual perception. A number of disorders are associated with psychotic symptoms, including schizophrenia (SZ), schizoaffective disorder (SA), and bipolar disorder (BP). Past studies have suggested that patients with psychosis demonstrate weaker surround suppression and reduced contrast sensitivity compared to controls. We tested for these effects by examining activity patterns across voxels with a decoding analysis (multi-voxel pattern analysis) as part of the ongoing Psychosis Human Connectome Project. High field (7 tesla) fMRI and behavioral data were collected during a visual contrast discrimination task. Small circular gratings were presented at four contrast levels (10, 20, 40, & 80%), each with and without a surrounding annular grating. FMRI responses were extracted from a retinotopically mapped region of primary visual cortex corresponding to the target gratings. We used support vector machine decoding with 5-fold cross validation to investigate differences in visual processing between people with psychosis, their first-degree relatives, and healthy controls. We expected less accurate decoding of conditions with and without surrounding stimuli, and lower decoding accuracy for contrast in people with psychosis compared to controls. Decoding accuracy rose above chance levels following stimulus presentation and declined following stimulus offset. Preliminary results show peak decoding accuracy for the presence versus absence of a surround was significantly higher in controls compared to participants with psychosis, consistent with our first hypothesis. However, we did not observe a significant difference in peak decoding accuracy between groups when decoding stimulus contrast. Our results suggest that psychosis is related to a reduction in the fidelity of information about surrounding stimuli in primary visual cortex.