Characterizing neural mechanisms of cognitive control

描述（由申请人提供）：我们研究项目的总体目的是了解认知控制的大脑机制，即根据内部目标灵活调整思想和行为的能力。前一个授权期的工作集中在信息处理中的冲突（例如，在繁忙的十字路口如何最好地驶入预定车道的不确定性）如何导致立即将注意力重新集中在手头的任务上（冲突控制）。在这个更新申请中，我们提出了一个宽泛的问题：下次你走近那个十字路口时会发生什么？你会记得你以前的困难，并以高度的注意力来处理它吗？认知控制和记忆过程之间的这种相互作用在我们的日常生活中无处不在，不能适当地将情境需求与注意力状态相匹配是许多使人衰弱的精神疾病（如精神分裂症）的一个关键特征。然而，线索（例如，交集）和适当的控制状态之间的联系是如何形成的，以及不同的控制操作对后续记忆的后果是什么，目前知之甚少，部分原因是控制在历史上被定义为与良好的学习反应（“意志与习惯”）相反。因此，本提案的目标是表征认知控制和联想过程的相互作用，以改进当前大脑如何支持适应性行为的模型，并为理解临床领域的认知控制失败提供新的方法。为此，我们从两个方向研究控制-记忆的相互作用：首先，具体目标1研究学习如何驱动认知控制的分配。换句话说，繁忙的十字路口是如何与高度控制状态联系在一起的？我们使用计算建模、fMRI和fMRI引导的TMS方法来描述大脑如何学习适应不断变化的需求的控制设置（研究1），并直接对比连接线索到控制状态（情境控制学习）的神经机制与支持经典刺激-刺激和刺激-反应学习的神经机制（研究2和3）。第二，具体目标2描述了认知控制过程如何影响记忆。例如，在你必须克服困难之后，你是否更有可能记住十字路口的关键细节？具体来说，我们将评估三种关键的认知控制操作（冲突控制、更新和反应抑制）对记忆编码的影响，方法是让参与者在不同的刺激下执行这些操作，然后在惊喜记忆测试中测试他们对这些刺激的识别能力（研究4-6）。将这种方法与功能磁共振成像（fMRI）相结合，通过对比任务相关和任务无关刺激编码成功（随后被记住）与失败（随后被遗忘）的神经特征，我们可以解析将特定控制操作与记忆联系起来的大脑机制。这个创新的项目将大大提高我们对人类大脑如何促进上下文敏感的、受控制的行为的理解，并深入了解这种能力是如何失败的。

DESCRIPTION (provided by applicant): The overall purpose of our research program is to understand the brain mechanisms of cognitive control, the ability to flexibly adapt thoughts and behavior in line with internal goals. Work in the previous grant period has focused on how conflicts in information processing (e.g., uncertainty over how to best steer into your intended lane at a busy intersection) can lead to an immediate re-focusing of attention on the task at hand (conflict- control). In this renewal application, we ask broadly: what happens the next time you approach that intersection? Will you remember your previous difficulties and approach it with a heightened focus of attention to start with? This type of interaction between cognitive control and memory processes pervades our daily lives, and a failure to appropriately match situational demands with attentional states is a key feature of many debilitating psychiatric disorders, such as schizophrenia. However, how associations are formed between cues (e.g., the intersection) and appropriate control states, and what the consequences are of different control operations for subsequent memory, is presently poorly understood, owing in part to control being defined historically in opposition to well-learned responses ("will vs. habit"). Therefore, the goal of this proposal is to characterize the interaction of cognitive control and associative processes in order to improve current models of how the brain supports adaptive behavior, and to enable new approaches for understanding failures of cognitive control in the clinical domain. To this end, we examine control-memory interactions from two directions: first, specific aim 1 investigates how learning drives the allocation of cognitive control. In other words how does the busy intersection become associated with a heightened control state? We use computational modeling, fMRI, and fMRI-guided TMS approaches to characterize how the brain learns to adapt control settings to changing demands (study 1), and to directly contrast the neural mechanisms that link cues to control states (context-control learning) with those supporting classic stimulus-stimulus and stimulus- response learning (studies 2 and 3). Second, specific aim 2 characterizes how cognitive control processes affect memory. E.g., are you more likely to remember key details about the intersection after you had to overcome difficulty there? Specifically, we will assess the impact of three crucial cognitive control operations (conflict-control, updating, and response inhibition) on memory encoding, by having participants perform these operations on different stimuli and subsequently testing their recognition of these stimuli i surprise memory tests (studies 4-6). Pairing this approach with fMRI allows us to parse the brain mechanisms that link specific control operations to memory by contrasting the neural signatures of successful (subsequently remembered) vs. unsuccessful (subsequently forgotten) encoding of task-relevant and task-irrelevant stimuli as a function of control state. This innovative projec is set to significantly improve our understanding of how the human brain facilitates context-sensitive, controlled behavior, and to gain insight into how this ability may fail.