Schizophrenia patients show augmented spatial frame illusion for visual and visuomotor tasks

Abstract

Previous research has identified several key processes of visual perception and visually guided action that are implicated in schizophrenia. Yet, it is not well understood whether similar or different brain mechanisms mediate the abnormalities in these two processes. To explore this issue, we examined visual and visuomotor processing in schizophrenia, utilizing an illusion known as the Roelofs effect. This illusion refers to the spatial mislocalization of an object within an off-centered frame, with the object appearing to be shifted towards the opposite direction of the frame offset. In this study, localization of the object was measured either by a direct visual response or by an immediate or delayed visuomotor (reaching-to-touch) response. Patients demonstrated significantly greater magnitudes of the Roelofs effect in all response modes, indicating the existence of excessive spatial contextual effects of the frame during the processing of visual and visuomotor information, and when the two types of information are integrated over a delayed visuomotor response condition. These results provide evidence for a hypothesis of improper inhibitory control as a common mechanism underpinning abnormal visual and visuomotor processes in this mental disorder.

Figure 1

Illustration of stimuli used to induce Roelofs effect. The top, middle and bottom panels are for three frame positions (left-shifted, center and right-shifted). In each panel, the target can appear in one of the five positions (far left, left, center, right or far right). The outer boundary represents the computer screen on which the frame and the target were displayed. Identical stimuli were used for visual, immediate and delayed visually-guided actions.

Figure 2

Averaged perceived or touched positions for the visual, immediate and delayed visuomotor task conditions. In each panel, the x-axis represents target position whereas the y-axis represents subjects’ responses. The dotted line indicates veridical responses to the target positions (i.e. no Roelofs effect). The filled and open symbols are for controls and patients, respectively. Error bars represent ±1 standard errors. For the visual task (bottom row), subjects’ responses are determined by converting the data derived from the five-alternative choices to corresponding target positions. For the visuomotor (top row) and delayed visuomotor (middle row) tasks, subjects’ responses are finger-touched positions on the touch-sensitive screen.

Figure 3

Summary of group responses for all task conditions. The left panel is for the visual, immediate and delayed visuomotor tasks, whereas the right panel is for the contrast detection task. In each panel, the x-axis represents groups and tasks, whereas the y-axis represents the Roelofs effects, defined as the differences between subjects’ responses under the central frame condition and under the left- or right-shifted frame condition. The filled and open bars are for patients and controls, respectively. Error bars represent ±1 standard errors.