Integrated brain network and cell-circuit models of slow network fluctuations

摘要：项目4的总体目标归入中心目标3：开发迭代 建模和实证研究之间的相互作用，以整合跨数据范围的知识。去做 因此，项目4将开发神经电路动力学的新计算模型，并将其应用于特征拟合 项目1-3中的多模式神经记录。模型将被用来检验假说并洞察 慢脑网络波动的动力学和生物物理机制及其影响 关于感觉信息的局部回路处理。目标1将适合一个动力系统模型来捕捉 由LFP、EEG和iEEG测量的皮质区域的光谱状态的动力学。此外，这些 区域模型将在大规模网络模型中相互连接，以模拟全脑的动力学 功能磁共振成像测量的现象，如功能连接性和CAP状态。我们将测试特定的 光谱状态中缓慢(约0.1-1赫兹)波动的假设可以通过双稳来捕捉 由噪音和适应引起的转变，甚至更慢的波动(例如，在唤醒或内部与 外部注意力)可以通过双稳态和单稳态动力体制之间的转换来捕捉，并且 这些动力学可以解释在fMRI中观察到的时空效应。目标2将在生物物理上发展 具有层流分辨率的新皮质电路的详细模型，专为带来大规模人类而设计 从iEEG/EEG到微尺度的细胞和电路级现象(细胞尖峰，LFP/CSD)。详细型号将为 应用于研究少数关键电路中的慢波动机制，这些机制是 项目3中的NHP。我们将系统地探索特定于细胞类型的属性和层 特定的丘脑皮质和皮质连接必须结合起来才能复制多尺度动力学 由NHP研究揭示。我们将检验特定的假设，即外生驱动的模式与 通道电导的细胞类型特异性神经调节可以诱导电路活动的缓慢波动， 从细胞活动到脑电，跨越电生理尺度和物种。我们还将 描述持续的缓慢波动如何影响对自下而上的感觉诱发信号的电路响应， 将缓慢的神经动力学与任务表现联系起来。探索性目标3将开发一个多尺度模型来 探索微电路和大规模网络动力学之间的相互作用。具体地说，我们将嵌入 来自Aim 2的生物物理详细微电路模型作为大规模网络中的不同节点，其中 其他节点被模拟为目标1中的唯象动力学系统。总的来说，目标 项目4将综合项目1-3中的多模式录音，以开发多尺度机械 大脑皮层动力学和SBNFS的计算模型。

ABSTRACT: The overarching goal of Project 4 is subsumed under Center Aim 3: Develop iterative interactions between modeling and empirical studies to integrate knowledge across data scales. To do so, Project 4 will develop novel computational models of neural circuit dynamics and apply them to fit features of multi-modal neural recordings in Projects 1-3. Models will be used to test hypotheses and gain insight into dynamical and biophysical mechanisms underlying slow brain network fluctuations (SBNFs) and their impact on local circuit processing of sensory information. Aim 1 will fit a dynamical systems model to capture the dynamics of spectral states in a cortical region, as measured by LFP, EEG, and iEEG. In addition, these regional models will be interconnected in a large-scale network model to simulate brain wide dynamical phenomena as measured by fMRI, such as functional connectivity and CAP states. We will test the specific hypotheses that slow (~0.1-1 Hz) fluctuations in spectral state can be captured through bistability with transitions induced by noise and adaptation, that even slower fluctuations (e.g, in arousal, or internal vs. external attention) can be captured by shifts between bistable and monostable dynamical regimes, and that these dynamics can account for spatiotemporal effects observed in fMRI. Aim 2 will develop biophysically detailed models of neocortical circuits with laminar resolution specifically designed to bring macroscale human iEEG/EEG to microscale cellular and circuit-level phenomena (cell spiking, LFP/CSD). Detailed models will be applied to study mechanisms of slow fluctuations in a small number key circuits that are the target of study in NHP in Project 3. We will systematically explore the manner in which cell type-specific properties, and layer specific thalamocortical and cortical connectivity, must be combined to replicate the multiscale dynamics revealed by NHP studies. We will test the specific hypothesis that patterns of exogenous drive together with cell-type-specific neuromodulation of channel conductances can induce slow fluctuations in circuit activity that translates across electrophysiological scales and species from cell activity up to EEG. We will also characterize how ongoing slow fluctuations impact circuit responses to bottom-up sensory evoked signals, linking slow neural dynamics to task performance. Exploratory Aim 3 will develop a multi-scale model to explore the interplay between microcircuit and large-scale network dynamics. Specifically, we will embed the biophysically detailed microcircuit models from Aim 2 as distinct nodes in a large-scale network in which the other nodes are simulated as phenomenological dynamical systems from Aim 1. Collectively, the aims of Project 4 will synthesize multi-modal recordings from Project 1-3 to develop multi-scale mechanistic computational models of cortical dynamics and SBNFs.