S41. INFLUENCE OF PRIORS ON MOTION PERCEPTION IN SCHIZOPHRENIA. (9th April 2019)

• Record Type: Journal Article
• Authors: Bansal, Sonia; Bae, Gi-Yeul; Robinson, Benjamin R; Luck, Steven J; Gold, James
• Is Part Of: Schizophrenia bulletin. Volume 45(2019)Supplement 2
• Page Start: S321
• Page End: S322
• Publication Date: 2019-04-09
• Subjects: Schizophrenia -- Periodicals; Schizophrenia -- Research -- Periodicals; 616.898005
• DOI: 10.1093/schbul/sbz020.586
• Languages: English
• ISSNs: 0586-7614

Abstract

Abstract: Background: An emerging model posits that perceptual inference deficits could lead to distorted internal models of the world, which could then explain the existence of abnormal beliefs or delusions experienced by people with schizophrenia(PSZ). Here, we used motion perception tasks to explore how construction of internal models of the environment are utilized unconsciously and automatically in uncertain circumstances to predict and disambiguate perceptual inputs and guide decisions in schizophrenia. Methods: We used a random dot kinematogram algorithm developed by Roitman & Shadlen (2002). In the first experiment (PSZ, N=47, Controls, N=38), we used 4 (20–100%) motion coherence levels and 2 motion durations (500 and 1000 ms). On each trial, direction of motion was selected randomly from 0.1 - 360°. Second, to test perceptual inference when motion direction changes, we implemented a similar task (PSZ, N=44, Controls=36) using 2 coherence levels (35% and 100%) in which, midway through the trials, motion direction changed by 90°. In both tasks, participants were to report the direction perceived at the end of the trial (the changed direction for the second task) using a mouse to align a bar onto the perceived direction of motion. Performance was quantified as the angular difference between the true direction and the reported direction. We quantified response errors using a three-component mixture model, assuming that participants' direction reports were a combination of 1) correct perception of motion direction with some precision, 2) random guesses, and 3) opposite direction perception of the true motion direction. Further, we derived response bias estimates based on (i) influence of the prior trial when the current trial was of low coherence and (ii) initial motion direction (in the motion change experiment). Results: We observed that in both tasks, precision estimates decreased with reducing coherence, and more so in PSZ, and guess rate estimates were higher in PSZ across all coherence levels. Importantly, we found that when the current trial was of a lower coherence level than the prior trial, the perceived motion direction in PSZ was influenced by the prior trial motion direction, as evidenced by a significantly greater response bias towards the prior than in controls. Further, in PSZ, this response bias correlated with mean precision estimates across low coherence trials. In the changed motion estimation task, we observed a similar pattern, whereby PSZ manifested lower precision estimates, and more so on low coherence trials, and this was associated with response bias towards initial motion direction. Interestingly, in PSZ, response bias measures (towards the prior) were associated with the preoccupation and conviction dimensions on the Peters Delusions Inventory, while precision measures were associated with visual hallucinations. Discussion: These results suggest that, in situations of uncertainty (in this case, low coherence trials), PSZ are more influenced by prior expectation and sensory evidence from the previous trial to discriminate new perceptual inputs and to guide decisions. These data also provide further evidence for the theoretical link between positive symptoms such as hallucinations and delusions and aberrant perceptual inference.