Abstract
Deficits in visual working memory (VWM) are well recognized in Schizophrenia (SCZ). However, the mechanisms associated with such impairments are unclear. Previous findings show that sensory processing may be involved in encoding stimulus features in VWM (Pasternak & Greenlee, 2005). Using an equivalent noise paradigm, we characterized the sources of noise that influence perceptual inefficiency, namely, internal noise and external noise filtering, and tested the hypothesis that noisier perceptual processing may be linked to impaired VWM in SCZ. METHODS: Participants (19 individuals with SCZ and 19 demographically matched neurotypical controls) performed two behavioral tasks, measuring sources of perceptual inefficiency and VWM precision. In the orientation discrimination task, participants judged the orientation of gratings (1 cycle/°; 1° radius; tilted ±45° from vertical) embedded in varying levels of external noise (0-21%). Contrast thresholds were measured, and modeled using the Perceptual Template Model (Lu & Dosher, 2008) to estimate the levels of internal noise and external noise filtering. In the VWM task, participants reproduced orientations of analogous gratings at 1s delay across different set sizes (1-4). VWM precision was estimated by fitting circular Gaussians on orientation errors. RESULTS: Individuals with SCZ showed increased internal noise and worse external noise filtering compared to controls. They also exhibited considerably worse VWM precision across all set sizes. Importantly, external noise filtering predicted VWM precision in both groups. Internal noise was negatively correlated with VWM precision only in controls. CONCLUSION: We show evidence that perceptual inefficiencies may impact VWM precision in both neurotypical individuals and those with SCZ. The results further demonstrate noisier visual perception in SCZ. This provides broad implications, as noisy perceptual processing may play an important role in higher order cognitive and social deficits associated with SCZ.
Meeting abstract presented at VSS 2017