Entropy-driven visual choice in Parkinson's disease

Abnormalities in visual search and navigational tasks were recognized in Parkinson's Disease (PD), mainly related to top-down mechanisms. The aim of the present study was to assess the possible independent effects of bottom-up mechanisms on visual exploration and perceptual decision-making in PD by using synthetic entropy-based visual inputs, in order to minimize their semantic value and possible top-down biases. Binocular eye movements were recorded at 1000 Hz using the Eyelink 1000 Plus system. Participants were shown pairs of grey-scale images with varying entropic properties and they were asked to choose between them while their gaze patterns were recorded. Latency of first saccade and choice together with entropy of both image at first saccade and chosen image were extracted. Data were obtained from both PD subjects and Healthy Controls (HCs). An age-related tendency towards high entropic choice together with a first saccade direction towards high entropic images were observed in PD subjects as compared to HCs. These two parameters resulted directly correlated. A direct correlation between latency of choice and disease severity was also observed among PD subjects. Results support the role of the visual scene's intrinsic entropic value in visual exploration and choice. Bottom-up mechanisms regulating visual and motor choice seem related to bradykinesia in PD subjects. An overall increase in choosing and fixating high entropic figures as age increases was found in PD with respect to HCs, indicating a progressive pathological basal ganglia dysfunction in mechanisms controlling visual-attention and motor targeting of high-entropic images.