EEG Estimation of Cortical Connectivity

描述（由申请人提供）：迫切需要开发和评估一种新的信号处理方法，该方法可以在非侵入性脑电图（EEG）数据的相对低信噪比下表征大脑区域之间的因果影响。该项目的长期目标是开发一套新的信号处理工具，用于研究皮层中时间动态网络的相互作用，通过非侵入性测量头皮的电或磁电位。这个R21应用程序的目标，是实现这一长期目标的第一步，是开发和评估一个新的因果影响框架，该框架集成了分布式解剖皮质源和多变量自回归（MVAR）时间动态网络模型的建模。成功完成这一目标将克服严重限制脑电数据皮层网络研究的技术差距，并展示集成源/网络建模的潜力。为了实现这一目标，将追求以下具体目标：1)开发一个综合框架，用于从头皮测量中估计有效的皮层连通性；2)评估人类在三种行为状态下自发脑电图记录的有效连通性。人体研究旨在验证以下假设：在看电影（醒着）时自下而上的信息流占主导地位，而在做梦时快速眼动睡眠（REM）时自上而下的信息流占主导地位。所提出的工作是创新的，因为(i)采用先进的信号处理方法和脑电图源模型直接从头皮记录推断皮层连接，（ii）实验数据很可能显示状态之间连接的变化，以及（iii）交叉验证评估策略将提供确定性的，定量的有效性证据。预计该项目的结果是：(i)在实际信噪比下评估皮层连通性的非侵入性框架；（ii）使用实验数据评估皮质网络模型质量的交叉验证方法；（三）对意识和睡眠本质的新见解。这项工作意义重大，因为开发一种可靠的、无创的评估皮层区域之间有效连接的工具，将使神经科学家能够解决许多关于正常大脑功能、发育和疾病的长期和新的假设。

DESCRIPTION (provided by applicant): There is a critical need to develop and evaluate a new signal processing approach that can characterize causal influences between brain regions at the relatively low SNRs typical of noninvasive electroencephalography (EEG) data. The long-term goal of this project is to develop a new set of signal processing tools for study of temporally dynamic network interactions in the cortex from noninvasive measurements of electric or magnetic potentials at the scalp. The objective of this R21 application, the first step toward achieving this long term goal, is to develop and evaluate a novel causal influence framework that integrates the modeling of distributed anatomical cortical sources and multivariate autoregressive (MVAR) temporally dynamic network models. Successful completion of this objective will overcome technical gaps that have severely limited study of cortical networks from EEG data and demonstrate the potential of integrated source/network modeling. The following specific aims will be pursued to achieve this objective: 1) Develop an integrated framework for estimating effective cortical connectivity from scalp measurements; and 2) Evaluate effective connectivity from spontaneous EEG recordings in humans during three behavioral states. The human studies are designed to test the hypothesis that bottom-up information flow is dominant while watching a movie (waking) and top-down information flow is dominant in REM sleep while dreaming. The proposed work is innovative because (i) advanced signal processing methods and EEG source models are employed to infer cortical connectivity directly from scalp recordings, (ii) the experimental data is very likely to exhibit changes in connectivity between states, and (iii) a cross-validation evaluation strategy will provide definitive, quantitative evidence of effectiveness. The outcomes of this project are expected to be: (i) a noninvasive framework for assessing cortical connectivity at realistic SNRs; (ii) a cross-validation methodology for evaluating the quality of cortical network models using experimental data; and (iii) new insights into the nature of consciousness and sleep. This work is significant because development of a tool for reliable, noninvasive assessment of effective connectivity between cortical regions will enable neuroscientists to address many longstanding and new hypotheses about normal brain function, development, and disease. PUBLIC HEALTH RELEVANCE: Contemporary neuroscience has established that human brain function arises from the time-varying neural activity of spatially distributed neural circuits in the cortex. This project develops a new set of signal processing tools for noninvasive interrogation of causal connectivity between different cortical regions and will further understanding of the brain in both health and disease.