Neural mechanisms underlying simple and complex reinforcement learning

描述（由申请人提供）：为了引导动机行为，人类和动物需要能够灵活地评估环境中刺激的情感价值-积极或消极。这一过程在精神疾病中可能会出错；例如，在创伤后应激障碍（PTSD）中，与创伤相关的某些刺激可能被赋予过于消极、不灵活的价值，即使在它们本不应具有威胁性的环境中也是如此。这个建议的重点是刺激的价值是如何在大脑中表现出来的，以及这种表现是如何随着环境的变化而变化的。为了研究这一点，当猴子执行两项任务中的一项时，将同时记录两个大脑区域——杏仁核和眶额皮质（OFC）的神经活动：一项是简单的条件反射任务，其中视觉刺激与奖励或惩罚联系在一起；另一项是情境依赖任务，其中视觉刺激的意义随上下文线索而随时变化。更简单的任务将揭示刺激值是否以及如何由OFC和杏仁核的单个神经元表示，并将提供有关这些“价值信号”的相对时间的信息，从而可能阐明这两个区域如何相互作用。更复杂的任务将揭示这些信号是否能够灵活地变化，以反映刺激的价值，当它可以根据环境快速变化时。它还可能揭示上下文信息——即解释刺激值的“规则”——是否以及如何在OFC和/或杏仁核中编码。总之，这些实验可能阐明了价值分配是如何形成的，然后由一个涉及边缘区域（如杏仁核）和可能涉及情绪反应执行控制的区域（如OFC）的神经回路来调节。了解这个回路如何正常运作是理解它如何在PTSD、临床抑郁症和精神分裂症等精神疾病中起作用的重要的第一步。相关性：为了做出日常决策，人（和动物）需要能够为环境中的对象分配值（例如“好”或“坏”），并且有时需要根据对象发生的上下文更改这些值。这一过程在许多精神疾病中被破坏，比如抑郁症和创伤后应激障碍。这一建议的重点是理解情感和认知大脑区域如何相互作用，以学习和修改对象的价值-这些知识可能有助于我们理解并最终治疗精神疾病的原因。

DESCRIPTION (provided by applicant): In order to guide motivated behavior, humans and animals need to be able to flexibly evaluate the affective values - positive or negative - of stimuli in the environment. This process may go awry in psychiatric disorders; for example, in post-traumatic stress disorder (PTSD), an overly negative, inflexible value may be assigned to certain stimuli associated a trauma, even in a context where they should be non- threatening. The focus of this proposal is on how the values of stimuli are represented in the brain, and how this representation changes in response to changing environmental context. To investigate this, neural activity will be recorded simultaneously in two brain areas - the amygdala and orbitofrontal cortex (OFC) - while monkeys perform one of two tasks: a simple conditioning task, in which visual stimuli become linked with reward or punishment, and a context-dependent task, in which the meaning of visual stimuli changes from moment to moment depending upon contextual cues. The simpler task will reveal whether and how stimulus value is represented by individual neurons in OFC and amygdala, and will provide information about the relative timing of these "value signals" that may clarify how these two areas interact. The more complex task will reveal whether these signals can change flexibly enough to reflect the value of a stimulus when it can change rapidly depending on context. It may also reveal whether and how contextual information - i.e., the "rule" for interpreting the value of a stimulus - is encoded in OFC and/or amygdala. Together, these experiments may elucidate how value assignments are formed and then modulated by a neural circuit involving limbic areas (such as the amygdala) and areas that may be involved in the executive control of emotional responses (such as OFC). Learning how this circuit functions normally is an important first step in understanding how it may dysfunction in psychiatric disorders like PTSD, clinical depression, and schizophrenia. Relevance: In order to make everyday decisions, people (and animals) need to be able to assign values (such as "good" or "bad") to objects in the environment, and sometimes to change these values depending on the context in which an object occurs. This process is disrupted in many psychiatric disorders, such as depression and post-traumatic stress disorder. The focus of this proposal is on understanding how emotional and cognitive brain areas interact to learn and modify the values of objects - knowledge that may aid our understanding, and eventually treatment, of the causes of psychiatric disorders.