Neurocognitive mechanisms of control over cognitive stability and flexibility

项目摘要/摘要 人类有一种独特的能力，能够将内在目标强加给他们如何与环境互动。 这种能力被称为“认知控制”，包括两个核心部分：(1)集中注意力的能力 当前与目标相关的刺激特征和反应(任务集)，而忽略与任务无关的特征(认知 稳定性)；以及(2)当环境变化时切换到不同任务集的能力(认知灵活性)。至关重要的是， 为了在动态的环境中茁壮成长，我们需要不断调整我们的认知稳定性和灵活性水平，以适应 不断变化的需求。例如，在做饭时，需要稳定(例如，切洋葱时要有很强的工作重点)和 灵活性(例如，在配方阅读和炉顶监测之间快速切换)随着时间的推移而频繁变化。这个 因此，对稳定性和灵活性的战略性监管是日常生活中取得成功的基础，实际上也是严重的 在许多临床情况下受损。然而，人们对潜在的神经认知机制知之甚少。这是 由于这样一个事实，虽然有大量关于认知稳定性的文献(以冲突控制研究的形式)和 灵活性(以任务转换研究的形式)，这些过程要么被孤立地研究，要么被合并， 或者没有在它们的动态适应方面被审问。本提案旨在克服这些障碍，以 通过结合一种新的任务方案取得进展，该任务方案评估同时和独立的稳定性和 灵活的计算建模、功能磁共振成像(FMRI)和颅内电信号 脑电图(IEEG)。我们的总体目标是描述并发的、策略性的 控制认知的稳定性和灵活性。我们通过三个目标三角化这个目标：目标1寻求建立第一个目标 基于行为数据的并发稳定性和灵活性调节计算模型 对稳定性和灵活性的时变要求的协议(研究1和2)。我们的工作模式包括两个 独立强化学习者对即将到来的稳定性需求进行逐一预测(冲突- 似然性)和灵活性(切换似然性)，这进而调制了不同的试验内漂移-扩散模型 参数。目标2使用致胜模型来确定调节这些调节的神经机制 稳定性和灵活性。在大量先前文献的基础上，我们使用互补的功能磁共振成像(研究3)和iEEG(研究4) 检验关于外侧前额叶、后叶各自作用的特定神经解剖学假说的方法 顶叶、前扣带回皮质以及基底节在支持稳定性的主动适应中的作用 以及对时变需求的灵活性。最后，Aim 3将使用功能磁共振成像来描述上下文的神经恢复- 当它们被反应性地应用时，即响应特定需求时，适当的稳定性和灵活性设置- 预测刺激(研究5)。总而言之，这些互补的目标代表着对 同时调节认知稳定性和灵活性的计算和神经机制。这 这一创新项目将极大地促进我们对认知控制的神经计算基础的理解， 为在临床条件下识别稳定性和灵活性调节的潜在失效模式奠定了基础。

PROJECT SUMMARY/ABSTRACT Humans have a uniquely developed ability to impose internal goals on how they interact with their environment. Referred to as “cognitive control”, this capacity includes two core components: (1) the ability to focus attention on currently goal-relevant stimulus features and responses (a “task set”) while ignoring task-irrelevant features (cognitive stability); and (2) the ability to switch to a different task set when circumstances change (cognitive flexibility). Crucially, to thrive in a dynamic environment, we need to continuously adapt our levels of cognitive stability and flexibility to suit changing demands. E.g., when cooking a meal, needs for stability (e.g., a strong task-focus when slicing onion) and flexibility (e.g., rapid shifting between recipe reading and stovetop monitoring) change frequently over time. The strategic regulation of stability and flexibility is thus fundamental for success in everyday life, and is in fact severely impaired in many clinical conditions. However, the underlying neurocognitive mechanisms are poorly understood. This is due to the fact that, while there are large literatures on cognitive stability (in the shape of conflict-control studies) and flexibility (in the shape of task-switching studies), these processes have been either investigated in isolation, conflated, or not interrogated in terms of their dynamic adaptation. The present proposal seeks to overcome these barriers to progress by combining a novel task protocol that assesses simultaneous and independent adaptive shifts in stability and flexibility with computational modeling, functional magnetic resonance imaging (fMRI), and intracranial electro- encephalography (iEEG). Our overall goal is to characterize the neurocognitive mechanisms of concurrent, strategic control over cognitive stability and flexibility. We triangulate this goal via three aims: Aim 1 seeks to establish the first computational model of concurrent stability and flexibility regulation by fitting and simulating behavioral data from protocols with time-varying demands on stability and flexibility (Studies 1 and 2). Our working model consists of two independent reinforcement learners making trial-by-trial predictions about forthcoming demands on stability (conflict- likelihood) and flexibility (switch-likelihood), which in turn modulate distinct within-trial drift-diffusion model parameters. Aim 2 employs the winning model to determine the neural mechanisms mediating these adjustments in stability and flexibility. Building on a large prior literature, we use complementary fMRI (Study 3) and iEEG (Study 4) approaches to test specific neuroanatomical hypotheses about the respective roles of the lateral prefrontal, posterior parietal, and anterior cingulate cortex, as well as the basal ganglia, in supporting the proactive adaptation of stability and flexibility to time-varying demands. Finally, Aim 3 will use fMRI to characterize the neural reinstatement of context- appropriate stability and flexibility settings when they are applied reactively, i.e., in response to specific demand- predicting stimuli (Study 5). Together, these complementary aims represent the first systematic investigation into the computational and neural mechanisms underlying the concurrent regulation of cognitive stability and flexibility. This innovative project will significantly advance our understanding of the neurocomputational bases of cognitive control, and lay the groundwork for identifying potential failure modes of stability and flexibility regulation in clinical conditions.