Measuring and modeling large-scale electrophysiological brain dynamics: From neural mechanisms to cognitive processes

该研究计划的总体目标是推进电生理脑网络动力学与人类高阶认知过程之间关系的机械表征。特别是，我们的研究将继续探索振荡脑动力学在各种高阶认知过程中的作用。在上一个周期取得的进展的基础上，我们将通过部署预测模型和关注决策和意识中涉及的节奏过程来扩大我们的研究。这个多学科的研究计划将跨越三个互补的研究轴：I。认知神经科学研究轴：拟议的研究将调查本地和远程振荡同步在认知和行为中的功能作用。我们将针对一系列认知过程，在短期和中期，重点是决策和意识（包括睡眠，麻醉和有意识与无意识的刺激处理）。二.系统神经科学研究轴：我们将寻求通过弥合多层次观察之间的差距来促进我们对节奏和神经网络的空间，时间和频谱动力学的理解。为此，我们将使用各种成像方法分析大脑数据。更具体地说，我们将比较脑磁图（MEG）和头皮EEG的测量结果与颅内EEG，微电极记录或功能性磁共振成像（fMRI）的数据。三.神经AI轴：该提案的一个重要部分将致力于实现计算模型和数据驱动工具，以建立一方面行为/认知与另一方面脑电生理测量之间的机械联系。这一拟议的研究主体包括（i）大规模数据挖掘（例如监督和无监督数据分类），（ii）计算预测建模（包括基于深度学习的AI启发模型）和（iii）MEG/EEG分析的方法学发展。在NSERC Discovery支持的最后一个周期中，我在实验室中建立了多学科的专业知识，使我们现在能够将我们的研究向前迈出重要的一步，扩大我们对复杂大脑网络动力学的理解，以及使用尖端技术（包括人工智能启发的方法）将其与高阶认知联系起来。通过转向大脑编码模型，这项研究将寻求建立更普遍和更好的振荡大脑模式如何塑造认知的机制特征。此外，该计划的高度跨学科性质（结合神经科学，机器学习和神经成像）以及我们对EDI原则的坚定承诺将为HQP创造一个强大，多样化和包容性的培训环境。

The overarching goal of this research program is to advance the mechanistic characterization of the relationship between electrophysiological brain network dynamics and higher-order cognitive processes in humans. In particular, our research will continue to probe the role of oscillatory brain dynamics in a variety of higher-order cognitive processes. Building on progress achieved in the last cycle, we will expand our research by deploying predictive models and by focusing on rhythmic processes involved in decision-making and consciousness. This multidisciplinary research program will span three complementary research axes: I. Cognitive neuroscience research axis: The proposed research will investigate the functional role of local and long-range oscillatory synchronization in cognition and behavior. We will target a range of cognitive processes with, in the short- and mid-term, a focus on decision-making and consciousness (including sleep, anesthesia and conscious versus unconscious stimulus processing). II. Systems neuroscience research axis: We will seek to advance our understanding of the spatial, temporal and spectral dynamics of rhythmic and arrhythmic brain networks by bridging the gap between multiple levels of observation. To this end, we will analyze brain data using a wide array of imaging methods. More specifically, we will compare measurements from Magnetoencephalography (MEG) and scalp-EEG with data from intracranial EEG, microelectrode recordings or functional magnetic resonance imaging (fMRI). III. Neuro-AI axis: An important part of this proposal will be dedicated to the implementation of computational models and data-driven tools to establish mechanistic links between behavior/cognition on one hand, and electrophysiological brain measurements on the other. This proposed body of research includes (i) large-scale datamining (e.g. supervised and unsupervised data classification), (ii) computational predictive modeling (incl. AI-inspired models based on deep learning) and (iii) methodological developments for MEG/EEG analyses. The multidisciplinary expertise I have established in my lab over the last cycle of NSERC Discovery support, allows us now to take our research an important step further, expanding our understanding of complex brain network dynamics and their link to higher-order cognition using cutting-edge techniques, including AI-inspired methods. By moving towards brain encoding models, this research will seek to establish more generalizable and better mechanistic characterizations of how oscillatory brain patterns shape cognition. In addition, the highly interdisciplinary nature of this program (combining neuroscience, machine learning and neuroimaging) and our very strong commitment to EDI principles will create a strong, diverse and inclusive training environment for HQP.