Perceptual Decision Making Under Conditions of Visual Uncertainty

Overview

In this proposal the investigators have three Specific Aims using human patient populations as model systems; 1) identify a role for the Basal Ganglia (BG) in perceptual decision making; 2) determine whether the Basal Ganglia contribute to decision making under conditions of visual uncertainty; 3) determine whether the cerebellum plays a role in perceptual decision-making under conditions of visual uncertainty. The investigators designed experiments using healthy humans and humans with diseases known to affect the Basal Ganglia and the cerebellum, Parkinson's Disease, dystonia and non-dystonic cerebellar damage. With this approach the investigators will test the following hypotheses: 1) Patients with Parkinson's Disease and dystonia will have more difficulty than healthy controls making perceptual decisions when faced with sensory uncertainty; when sensory information is certain, patients will show improved decision-making but will still be impaired relative to healthy humans. Hypothesis 2: If ambiguous sensory information is aided by prior information, patients with Parkinson's Disease and dystonia will be unable to use the prior (bias/memory) information to inform their decisions. Hypothesis 3: Deep Brain Stimulation (DBS) of Basal Ganglia structures will improve the ability of patients to use prior information to inform their decisions when faced with sensory uncertainty. Hypothesis 4: Both cholinergic and dopaminergic medical therapies will improve the ability of patients to use prior information to inform their decisions. Hypothesis 5: Patients with non-dystonic cerebellar damage will be similar to healthy controls in performance of a perceptual decision making task in conditions of visual uncertainty. The overarching framework of this application is that the same mechanisms (D1 striatal synaptic plasticity) that operate in reward learning play a role in learning and using stimulus priors in a perceptual decision-making task when faced with uncertainty. Because Parkinson's Disease and dystonia share deficits in striatal circuitry, the patient deficits on this task will be similar. Because non-dystonic cerebellar patients do not have dysfunction of striatal circuits, they will show no deficits in the ability to use stimulus priors to guide choices in uncertain conditions. In the event these patients do show deficits, this is will provide evidence for an unexplored role for the cerebellum in perceptual decision-making.

In this study, the investigators aim determine whether patients with Parkinson's disease (PD) have altered perceptual decision-making. The results of these experiments will shed important light on some of the enigmatic symptoms seen in patients with disorders of movement. In this study the subjects perform a computer generated visual task which involves reporting the direction of an arrow that is more or less difficult to see on the computer screen. Eye movements will be monitored using a video eye tracker. The system is completely non-invasive and positioned in front, out of the subject's line of sight to avoid interfering with their ability to perform the task. Subjects will report their decision with either an eye movement in the same direction as the perception of the arrow direction. Trials will proceed as follows: a central point will appear on the screen. Subjects are asked to look at the fixation point. After this random delay time, a patch of spots will appear in place of the center fixation spot, embedded in the patch will be an arrow. The contrast between the arrow and the patch will vary from high values to low. At high contrasts the direction at which the arrow points will be very clear and at low contrasts it will be unclear. The task of the subject is to make a rapid eye movement as soon as they decide which direction the arrow points. A sound will be played to indicate a correct choice. On trials in which the sensory information is ambiguous subjects will be told to 'make their best guess' to which target the arrow points. Unknown to the subject the investigators will manipulate the probability that one of the targets is the correct one. This latter manipulation will test whether subjects can use probability information to inform their decisions. The investigators are particularly interested in whether this probability information is used preferentially when the sensory information is ambiguous. The total number of subjects to be enrolled in the study is approximately 120. Subject groups are as follows: Healthy humans between the ages of 30 and 80 and those with PD (between ages of 45 and 80), dystonia (ages 30-80) and cerebellar lesions (ages 30-80) including those who have undergone DBS surgery to treat their PD or dystonia. The total duration of each session is approximately 2 hours. Each patient group including those with DBS will perform two sessions of the same task - one while on medication and one while off medication, hence the total duration of the study will be 4 hours for these groups (2 different sessions during different days). These patient groups are free to do these sessions at the same time as their routine physician visits or any other time that works best for them.