Cortical Representations for Value-Based Decision Making

点击翻译按钮获取中文摘要

Making good decisions -- in domains such as diet, finances, educational attainment, and public policy -- requires evaluating and comparing options that may be complex and multifaceted. Previous brain imaging research has shown that a frontal region called the ventromedial prefrontal cortex (vmPFC) has an important role in producing subjective value estimates for individual options during decision making. Activity in this region varies with the subjective value of the options under consideration. However, vmPFC also participates in many other cognitive functions, and its internal functional organization is not yet understood in detail. In particular, it is not yet known whether valuation signals originate from the same parts of vmPFC that have been implicated in processes such as memory, planning, and self-referential thinking. The present research will compare the patterns of vmPFC activity associated with valuation and non-valuation processes in individual brains. The results will help develop improved methods for measuring the engagement of specific cognitive processes from brain imaging data. Because vmPFC is thought to encode variables akin to economic utility, this work will contribute to the growing interdisciplinary dialogue between neuroscience and economics. The project will enhance the usefulness of neuroscience methods for understanding economic decision making and thereby contribute to the advancement of societally relevant economic goals. The interdisciplinary nature of the project will create rich educational opportunities for students and trainees involved in conducting the research. The project will use functional magnetic resonance imaging (fMRI) to test whether valuation effects in vmPFC can be topographically segregated from default-mode-network (DMN) effects when examined at the level of the single individual. The project involves a collaboration between researchers with expertise in neuroeconomics and vision science, as a means to import high-fidelity data analysis techniques from visual neuroscience to studies of decision making. Valuation tasks will be interleaved with manipulations of DMN deactivation in a test-retest protocol. Valuation and DMN effects have been shown to be diffusely co-localized within vmPFC in existing meta-analyses, but it is unknown whether they originate from the same or different cortical parcels in individuals. A test-retest design will allow comparisons of activation patterns within and across tasks. One possible outcome is that valuation and DMN effects may turn out to be associated with segregable patterns of cortical engagement. Such a finding would drive the development of sensitive and specific methods for reading out valuation-related activity, and would provide new tools for studying how the brain's valuation system is modulated by learning, context, and the engagement of control processes. An alternative outcome is that valuation and DMN effects may exhibit high topographic correspondence. This outcome would motivate new theoretical models of shared neural operations between these seemingly disparate cognitive processes.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.