CRCNS-US-German research collaboration on functional neuro-poroelastography

描述（由申请人提供）：由施加的神经刺激引起的、fMRI捕获的诱发血流动力学反应是神经元活动的间接测量，已成为现代临床神经科学脑功能研究的主要手段。然而，相对于负责神经血管耦合的潜在分子/细胞机制的信号解释是一个活跃的研究领域，并且看似矛盾的结果继续出现在文献中。部分问题是缺乏直接评估封闭颅骨局部神经活动的非侵入性成像选择，而开发新的成像方法仍然是一个重大挑战。正如本项目所述，我们已经确定了一种新的可能性——功能性神经孔隙弹性成像（fNPE），它将大脑中脑血管脉动的MRI采集与计算方法相结合，以估计空间局部的机械和流体力学脑组织特性。fNPE捕捉脑组织的机械功能反应，并将基于机械特性的变化提供其活动的第一个空间地图，无创且无外源性头部刺激。fNPE对神经组织的多尺度机械网络敏感，因此将揭示感官信号如何与大脑结构适应相关联。这些基本的新信息可以添加到电激活、代谢和结构的神经计算模型中。我们将开发具有非线性反演的fNPE，以获得除剪切模量外的水力导电性、间隙压力和流体分数的3D图像。这些结果将与一种新的宽带fMRE方法进行比较，其中图像也将通过非线性反演形成，但具有粘弹性模型。我们将使用已建立的刺激方案来定义这些新的MRE方法与脑功能的关联。神经网络不仅在大脑内传递电信号，它还提供了许多维持结构和形状的机械支架。液体的流动性控制着触发神经元活动所需的血流量和离子梯度，但它也会减弱组织运动并影响头盖骨内的动脉、静脉和间质压力。因此，我们假设脑血管流和相关的组织力学特性有助于正常的脑功能，反之亦然——脑功能调节脑血流动力学，并随之调节脑组织力学。目前，对正常和病理条件下的体内脑力学的认识和理解有限；然而，专门研究大脑力学和流体力学特性的MRI方法，即MRE技术正在兴起，并且将这些特性与大脑功能联系起来的初步研究也开始出现在文献中。尽管最近取得了一些进展，但在脑MRE中神经计算模型的反演还不发达，特别是如果我们对脑功能与脑力学和流体力学特性之间关系的理解取得了根本性的进展，则需要加以阐明。MRE中的神经计算，包括流体力学、孔隙弹性和粘弹性网络，可以在神经元健康（和功能）的框架内开辟一个新的研究领域，将机械结构与脑组织功能联系起来。鉴于神经元网络是机械脑支架的主要贡献者，这些发展也将为神经变性模型提供信息。这个项目将巩固最近开始的神经计算成像的国际合作。它还将加速实现新的成像和计算方法，用于临床神经科学对人脑功能的研究，并创建一个适用于其他器官系统和疾病的计算成像框架。思想和专门知识的国际交流将加强参与机构的研究基础和有关研究人员的知识。机构和研究团队都将拥有神经计算反演算法以及在拟议资助期结束前进行fNPE研究所需的MRI序列。研究生和博士后研究员不仅将接受高级MRI和计算方法的培训，而且还将通过在每个机构驻留时间来参与多学科国际合作并从中受益。

DESCRIPTION (provided by applicant): Evoked hemodynamic response caused by an applied neurological stimulus and captured with fMRI is an indirect measure of neuronal activity that has become the work-horse of modern clinical neuroscience research on brain function. However, interpretation of the signal relative to the underlying molecular/cellular mechanisms responsible for the neurovascular coupling is an active area of investigation, and seemingly contradictory results continue to appear in the literature. Part of the problem is lack of noninvasive imaging options that directly assess local neural activity in the closed cranium, and development of new imaging approaches remains a significant challenge. As described in this project, we have identified a new possibility - functional neuro-poroelastography fNPE) which combines MRI acquisition of cerebrovascular pulsation in the brain with computational methods to estimate spatially localized mechanical and hydrodynamical brain tissue properties. fNPE captures mechano-functional responses of brain tissue and will provide the first spatial maps of its activity based on changes in mechanical properties, noninvasively and without exogenous head stimulation. fNPE is sensitive to multiscale mechanical networks of neural tissue, and thus, will reveal how sensory signals are linked to structural brain adaptation. This fundamentally new information can be added to neuro-computational models of electrical activation, metabolism and structure. We will develop fNPE with nonlinear inversion to yield 3D images of hydraulic conductivity, interstitial pressure and fluid fraction in addition to shear modulus. These results will be compared to a new wideband fMRE approach where images will also be formed through nonlinear inversion but with viscoelastic models. We will define the association of these new MRE methods with brain function using established stimulus protocols. The neuronal network not only transmits electrical signals within the brain, it also provides much of the mechanical scaffold which maintains structure and shape. The mobility of fluid controls the blood flow and ionic gradients required to trigger neuronal activity, but it also attenuates tissue motion and influences the arterial, venous and interstitial pressures within the cranium. Thus, we hypothesize that cerebrovascular flow and related tissue mechanical properties contribute to normal brain function and/or vice versa - brain function modulates cerebral hemodynamics, and concomitantly, brain tissue mechanics. Currently, limited knowledge and understanding exist on in vivo brain mechanics under normal and pathological conditions; yet, MRI methods specific to the mechanical and hydrodynamical properties of the brain, namely MRE techniques, are emerging and preliminary studies relating these properties to brain function are beginning to appear in the literature. Despite recent advances, the neurocomputational model inversion is under-developed in brain MRE, especially if fundamental advances in our understanding of the relationships between brain function and brain mechanical and hydrodynamical properties are to be elucidated. Neuro-computation in MRE, which includes fluid dynamics, poroelasticity and viscoelastic networks may open a new field of study within the framework of neuronal health (and function), which relates mechanical structure with brain tissue function. These developments will also inform models of neuro-degeneration given that the neuronal network is major contributor to the mechanical brain scaffold. This project will solidify an international collaboration in neuro-computational imaging that was recently begun. It will also accelerate realization of new imaging and computational methods for clinical neuroscience research on human brain function, as well as create a computational imaging framework that is applicable to other organ systems and diseases. The international exchange of ideas and expertise will strengthen the research base at the participating institutions and the knowledge of the investigators involved. Both institutions and research teams will have the neuro-computational inversion algorithms along with the MRI sequences required to deliver fNPE studies by the end of the proposed funding period. Graduate students and post-doctoral fellows will not only be trained in advanced MRI and computation methods, but they will also be exposed to and benefit from participating in a multi-disciplinary international collaboration by spending time-in-residenc at each institution.