Multimodal integration of functional neuroimaging data

基于不同的物理和生理原理，现在有几种非侵入性的神经成像技术可用于研究大脑的功能。然而，在毫米尺度和毫秒尺度上探索脑神经元活动的理想技术并不存在。因此，有必要整合多模式数据，以克服每一种模式的内在局限性，并加强它们的联合使用。 我们建议使用多模式整合来阐明驱动大脑活动的两个重要过程的基本基础： (I)神经血管偶联(NVC)描述了神经元生物电活动和脑血流动力学反应之间的关系，将所需的营养物质(氧气、葡萄糖)带到活动区。大多数流行的功能成像技术只监测这些间接的血流动力学反应，发生在神经元放电后几秒钟。 (Ii)功能连接模式测量血流动力学信号的缓慢波动(&lt；0.05赫兹)，以表征大脑活动的潜在静息状态网络(RSN)结构。这些模式是在空间中组织的，并且从一个主题到另一个主题具有高度的重复性。功能连接性提供了一种独特的视角，可以从仅几分钟的静息状态功能磁共振成像(FMRI)数据中识别基于RSNs的生物标记物。 目的1：在多模式研究中利用直接测量神经元活动的优势。在对脑电/脑磁图(EEG/MEG)源定位领域做出重大贡献后，我们建议优化并仔细验证静息状态下的EEG/MEG源定位，以理清RSNs持续的血流动力学波动的神经元起源。 目的2：从多模式静息状态数据中建立新的敏感和特异的生物标志物。使用我们开发的根据静息状态fMRI数据估计连接器中枢的新方法(Lee等人NeuroImage 2016)，我们将评估连接器中枢是否表现出特定的动态(脑电/脑磁源成像)和代谢(氧合和葡萄糖基线代谢)特征，为表征连接器中枢在RSN波动中的驱动作用提供了一个独特的框架。 目的3：开发个性化/便携式成像设备来研究日常生活中的脑活动。利用红外光在脑组织中的吸收特性，我们开发了个性化的脑电/近红外光谱(NIRS)研究，以监测大脑皮质氧合和脱氧血红蛋白的波动。我们建议开发个性化的整夜EEG/NIRS，首次绘制正常睡眠期间的血流动力学波动、睡眠障碍以及睡眠和癫痫活动之间的相互作用。 所有建议的方法都将在健康对照组和局灶性癫痫患者身上进行评估。

Based on different physical and physiological principles, several non-invasive neuroimaging techniques are now available to study the brain in function. However, the ideal technique that would explore brain neuronal activity at a millimeter scale and at a millisecond scale does not exist. Therefore, it is necessary to integrate multimodal data, to overcome the intrinsic limitations of each individual modality and to potentiate their combined use. We propose to use multimodal integration to elucidate the fundamental bases of two important processes driving brain activity: (i) Neurovascular coupling (NVC) describes the relationship between neuronal bioelectric activity and hemodynamic response of the brain bringing the required nutrients (oxygen, glucose) to the active region. Most popular functional imaging techniques are monitoring only these indirect hemodynamic responses, occurring several seconds after the neuronal discharge. (ii) Functional connectivity patterns measure slow fluctuations of hemodynamic signals (<0.05Hz) to characterize the underlying resting-state networks (RSNs) structure of brain activity. These patterns are organized in space and highly reproducible from subject to subject. Functional connectivity offers a unique perspective to identify RSNs-based biomarkers from only few minutes of resting state functional Magnetic Resonance Imaging (fMRI) data. Objective 1: Taking advantage from direct measurements of neuronal activity in multimodal studies. Having significantly contributed to the domain of Electro-/Magneto-EncephaloGraphy (EEG/MEG) source localization, we propose to optimize and carefully validate EEG/MEG source localization during resting state, to disentangle the neuronal origins of RSNs ongoing hemodynamic fluctuations. Objective 2: Establishing new sensitive and specific biomarkers from multimodal resting state data. Using a new method we developed to estimate connector hubs from resting state fMRI data (Lee et al Neuroimage 2016), we will assess if connector hubs exhibit specific dynamic (EEG/MEG source imaging) and metabolism (oxygenation and glucose baseline metabolism) signatures, providing a unique framework to characterize the driving role of connector hubs in RSN fluctuations. Objective 3: Developing personalized/portable imaging modalities to study brain activity in everyday life. Exploiting absorption properties of infrared light within brain tissues, we developed personalized EEG/Near InfraRed Spectroscopy (NIRS) investigations to monitor cortical fluctuations of oxy- and deoxy-hemoglobin. We propose to develop personalized EEG/NIRS during the whole night, to map for the first time, hemodynamic fluctuations during normal sleep, sleep disorders and interactions between sleep and epileptic activity. All proposed methods will be evaluated on healthy controls and patients with focal epilepsy.