Perceptual decision making relies on cognitive processes such as acquiring and integrating sensory information, holding information in working memory, incorporating biases, setting a speed-accuracy regime, and planning a motor response. Many of these processes are affected (compared to age-matched controls) in patients with early Alzheimer's disease (AD). The neural correlates of these cognitive processes have been identified in persistently active neurons in frontal and parietal association cortex. We will test the hypothesis that disruption of persistent activity underlies some of the deficits seen in decision making in AD.
Our first aim i s to characterize the ability of patients with AD to incorporate evidence, environmental biases, and time pressure into their decisions. We will leverage the insights into the neural and computational basis of these abilities obtained from a well-studied perceptual decision making task. By comparing the performance of patients in this task against age matched controls, we will gain insights into the nature of the neural computations that are disrupted in early AD. The experiments also have the potential to uncover new behavioral markers for early AD.
Our second aim i s to mimic the deficits seen in AD in the macaque monkey by manipulating persistent activity in parietal association cortex while they perform the same perceptual decision task. We will bilaterally express inhibitory chemogenetic DREADD receptors (Designed Receptors Exclusively Activated by Designer Drugs) in a subregion of parietal cortex with neurons that show persistent activity in this decision making task. Preliminary data shows that we can successfully change decision making behavior with this approach. We will build upon these results by investigating how integrating evidence, incorporating biases, and deciding under time pressure is affected by this manipulation in the same task as used with AD patients. Together, our results will provide insights into the computations that underlie decision making, their neural implementation in the primate brain, and how failure to sustain persistent activity in association cortex can lead to deficits in decision making in AD. Our long-term goal is to develop behavioral assays for early diagnosis and to gain insight into fundamental mechanisms that will ultimately lead to new therapeutic targets in AD.
We will combine psychophysics and computational modeling to characterize the neural computations affected in decision making in patients with Alzheimer's disease (AD). We will mimic the behavioral deficits seen in AD by disrupting persistent activity in parietal cortex of the macaque monkey while they perform the same decision making task. The proposed project will give us insights into the relationship between neural activity, neural computations, and behavioral deficits in AD.