Aster Perkins - US grants

Project Summary There is often not an obvious best choice when making a decision, particularly when deciding between options that have multiple features, or attributes. For example, when deciding between lunch options, one might factor the taste, price, and nutritional quality of an option into one's selection. Additionally, such choices may also be in?uenced by other processes, such as the choice context or one's attention. The orbitofrontal cortex (OFC) plays an important role in representing associations between stimuli and values, and has been suggested to represent the overall value of an option integrated across its attributes. However, it remains unknown whether options are necessarily evaluated only on the basis of an integrated value, as suboptimal decision-making e?ects like the attraction e?ect point to the possibility of within-attribute comparison. Furthermore, choice behavior is known to interact with visual attention to produce choice biases, indicating a role for the frontal eye ?elds (FEF), an area linked to visual attention and also known to respond to the value of visual targets. In order to investigate how multi-attribute choices are instantiated in neural activity, we will train two rhesus macaques on a multi-attribute decision making task, in which each of two simultaneously-presented options consists of two attributes: the sweetness of that option's sucrose reward, and the probability that the animal will receive that reward. The information about each attribute is represented by separate bars that either increase or decrease with increasing attribute value, a relationship that was denoted by the color of the bar. This allows us to investigate both free-viewing gaze behavior and changes in choice behavior due to di?erences in attribute value as well as visual presentation. Our preliminary data show that when the stimuli of shared attributes matched each other across option (e.g., both reward bars increased in size with increasing sweetness), choice behavior improved from baseline, while when the stimuli of shared attributes did not match each other across option (e.g., reward bar A increased in size with increasing sweetness, while reward bar B decreased), choice of the better option became suboptimal, implying a role for within-attribute comparison in the choice process. This proposal combines robust behavioral techniques with high channel-count multi-unit recording in OFC and FEF to investigate how multi- attribute options are compared and how attention is involved in complex decisions. We hypothesize that correlates of attribute value and comparison will be present in the OFC, and that functional connectivity between OFC and FEF will depend on the focus of visual attention during choice and valuation. Our ultimate goal is to further elucidate the neural mechanisms of complex decision-making to better guide the development of treatments for neuropsychiatric conditions in which it is impaired.