Abstract
In the last four decades, evidence has steadily amassed for the existence of visual abnormalities in Parkinson's Disease (PD), reflected in psychophysics and electrophysiology in both human PD patients and animal models. Visual perceptual deficits in contrast sensitivity have been accompanied by deficits in visual evoked responses in both the electroretinogram (ERG) and scalp visual evoked potentials (VEP), which have been partially attributed to dopamine depletion in the retina. Despite their central role in vision, the effects of PD on normalization or “gain-control” computations, which pervade visual processing, have not yet been examined. Here we used a novel steady-state visual evoked potential (ssVEP) paradigm in the form of a passive-viewing task, which allows direct measurement of gain control effects in space (surround suppression) and the corresponding time profiles (sensory adaptation). We found that PD patients exhibited greater suppression of ssVEP amplitude as a function of surrounding stimulus contrast, and had ssVEP responses that more steeply adapted over time. Furthermore, frequency spectrum in PD patients shows a significantly higher level of noise than control subjects. The profile of higher suppression by the stimulus context, steep adaptation over time, embedded within an abnormally greater background noise level might explain the contrast sensitivity abnormalities in PD, since an exaggerated gain control may contribute to the reduced visual contrast sensitivity often observed in PD patients. Results from a simple computational model that when fit to the ssVEP data could potentially provide valuable diagnostic measures.