Abstract
Several studies have reported that visual perception in schizophrenic patients, compared to healthy individuals, is relatively immune to illusions induced by surrounding visual context (Daki, Carlin, & Hemsley, 2005; Tadin et al., 2006; Uhlhaas et al., 2006). Given the important role of context in normal vision, these abnormalities might have considerable impact on visual processing in schizophrenia. Prior studies, however, focused on single visual sub-modalities (e.g., motion), precluding more general conclusions about contextual deficits in schizophrenia. In this project, we investigated contextual effects across a variety of visual tasks in normal adults and in adults diagnosed with schizophrenia or with bipolar disorder. We used adaptive psychophysical techniques to measure six different aspects of visual appearance of a central stimulus viewed in the presence of a surrounding visual context. In separate test runs, contextual modulations were measured using center-surround differences in luminance, orientation, contrast, size, or two different versions of motion direction. All participants, healthy adults and patients, performed accurately on control conditions in which each task was administered in the absence of surround context, verifying that participants understood and could perform the tasks. Healthy adults showed robust surround context effects on all tasks, as evidenced by pronounced misperception of the brightness, orientation, contrast, size, or motion direction of the center stimulus. Based on evidence to date, schizophrenic patients (N=8) were less influenced by contextual effects (meaning they performed more accurately than controls) induced by surrounding tilt, contrast, and motion. On some tasks individuals with bipolar disorder (N=11) produced results similar to those of schizophrenics, implying that neural processes underlying contextual interactions may be defective in clinical populations besides schizophrenics. Quantifying the strength and selectivity of visual contextual effects in schizophrenia could determine whether a common underlying deficit exists in contextual processing.