The role of causal network interactions in the retention of information in working memory: An empirical and theoretical investigation

描述(申请人提供)：工作记忆(WM)是一种核心的心理结构，在日常活动中发挥着核心作用，从相对简单的任务，如暂时保留与任务相关的信息，到在复杂的认知任务中操纵和使用这些信息，以及控制和指导适应行为。此外，个人在工作记忆中保留和操纵信息的能力已被证明是导致广泛的实验和现实世界测量中个体差异的一个重要因素，工作记忆功能障碍与精神疾病中观察到的认知缺陷密切相关，最明显的是精神分裂症。当前认知神经科学研究中的一个突出主题是，WM是由连接在一起形成功能网络的专门大脑区域之间的因果交互支持的。然而，支持这一提议的大部分证据来自对神经数据(fMRI、PET、EEG)的功能连通性分析，它给出了特定大脑区域的功能属性的相关指示，而不是因果指示，以及网络交互可能如何支持认知。此外，缺乏对因果网络相互作用如何产生口碑功能的因果和机械性描述。这项拟议的研究使用非侵入性脑刺激(TMS)和神经记录(EEG)来直接探索WM保留过程中的因果区域间交互作用，并通过神经合理的建模来提供这些交互作用如何随着认知状态变化而变化的机械描述。具体目的1使用TMS-EEG相结合的方法来阐明大脑前部和后部区域之间的因果相互作用在视觉信息短期保持中的作用。先前的工作表明，前额叶皮质(PFC)在自上而下控制后感觉区的活动中起着普遍的作用。尽管PFC已被证明通过过滤分心信息来调节WM的维持，但在没有分心的情况下，PFC的激活和前后部连接性的增加表明，这种相互作用可能也涉及纯粹的存储功能。这种可能性将通过刺激保留期间的额叶皮质来检验。 记录WM中的信息，并使用脑电记录刺激区域和远端皮质部位的反应。如果前额-后部功能交互在维持中起直接作用，我们预计连接的强度将随着WM中保存的信息量的增加而增加。同样，我们希望观察到，在需要在不同的后脑区编码的信息存储的任务的延迟期刺激PFC时，TMS诱发的激活到后脑区的空间扩散的系统性变化。具体目标2将使用现有的WM和变化检测的神经模型来实现系统级的连接模型，该模型可以捕获WM维护的脑电特征，并解释目标1中预期的TMS/EEG的观察到的状态依赖。建立神经建模和TMS/EEG之间的这种联系将为未来的工作奠定基础，旨在正式测试关于因果网络交互如何支持WM和其他认知功能的不同预测，并将为检查潜在的致衰性精神疾病症状(如精神分裂症)的潜在异常网络交互作用提供基础。

DESCRIPTION (provided by applicant): Working memory (WM) is a core psychological construct that plays a central role in everyday activities, from relatively simple tasks such as th temporary retention of task-relevant information, to the manipulation and use of this information in complex cognitive tasks and in the control and guidance of adaptive behaviors. Additionally, an individual's ability to retain and manipulate information in WM has been shown to be an important factor underlying individual differences across a broad spectrum of experimental and real world measures, and WM dysfunction has been strongly implicated in the cognitive deficits observed in psychiatric illnesses, most notably schizophrenia. A prominent theme in current cognitive neuroscience research is the idea that WM is supported by causal interactions between specialized brain areas linked together into functional networks. However, much of the evidence supporting this proposal has come from functional connectivity analysis of neural data (fMRI, PET, EEG), which gives a correlational rather than causal indication of the functional properties of particular brain areas, and of how network interactions might support cognition. Furthermore, a causal and mechanistic description of how causal network interactions give rise to WM functions is lacking. The proposed research addresses these issues using non-invasive brain stimulation (TMS), and neural recording (EEG) to directly probe causal interregional interactions during retention in WM, and neurally plausible modeling to provide a mechanistic description of how such interactions vary as a function of cognitive state changes. Specific Aim 1 uses combined TMS-EEG to clarify the role of causal interactions between frontal and posterior brain areas in the short-term retention of visual information. Prior work has suggested a general role for the prefrontal cortex (PFC) in the top-down control of activity in posterior sensory areas. Although the PFC has been shown to regulate WM maintenance by filtering out distracting information, PFC activation and increased frontal-posterior connectivity in the absence of distraction suggests that such interactions might also be involved in pure storage functions. This possibility will be examined by stimulating the frontal cortex during the retention of information in WM and recording the resulting response, both within the stimulated area and at distal cortical sites, using EEG. If frontal-posterior functional interactions play a direct rol in maintenance, we expect the strength of connectivity to increase with the amount of information being held in WM. Similarly, we expect to observe systematic changes in the spatial spread of TMS-evoked activations to posterior brain areas when the PFC is stimulated during the delay period of tasks requiring the storage of information that is coded in distinct posterior brain area. Specific Aim 2 will use an existing neural model of WM and change detection to implement a systems-level model of connectivity that can capture EEG signatures of WM maintenance and account for the observed state-dependence of TMS/EEG expected in Aim 1. Establishing this link between neural modeling and TMS/EEG will lay the groundwork for future work aimed at formally testing different predictions regarding how causal network interactions support WM and other cognitive functions, and will provide a basis for examining possible aberrant network interactions underlying the symptoms of debilitating mental illnesses, such as schizophrenia.