Dynamical complex network approaches for the analysis and modeling of large scale brain activities during cognitive processes

我们建议研究认知五个过程的神经网络基础。这将通过使用非线性动态复杂网络的原理来分析和建模大规模的事件相关脑反应(ERP)，并融合电生理和结构性神经成像数据来完成。在动态复杂网络分析和建模[洪堡大学-柏林大学非线性动力学小组]、生物心理学[洪堡大学心理学系]和神经科学的信号和信息处理[柏林洪堡大学计算机科学系]的专业知识之间的密切合作下，我们将追求以下目标：(I)从心理学实验中获得与阅读和(作为基准)相关的明确认知过程的高分辨率ERP数据。(Ii)提取传感器和源空间中功能和有效的(因果)连接，以监控认知过程中大脑活动的连续认知五个子过程。这些大规模的大脑网络将使用基于分层复杂网络概念的先进工具进行分析，以表征与不同实验条件相关的FIME演变连接性。(Iii)开发能够解释类似ERP活动的生物物理复杂动态网络模型，代表正在研究的过程中涉及的大脑区域的激活。这种模型的耦合参数将使我们能够识别处理流中的因果连通性。为了改进分析和建模，我们将：(A)获得参与者的结构MRT，以用于大脑活动的准确来源定位，以及(B)使用诸如US DTI和/或HARDI或DSI的扩散成像和足够的轨迹成像方法，将结构连通性信息整合到网络建模中，以估计参与者的大脑结构网络。

We propose to study neural network foundations of cognifive processes. This wilt be done by analyzing and modeling large-scale event-related brain responses (ERP) using the principles of nonlinear dynamical complex networks, and by fusioning of electrophysiological and structural neuroimaging data. In a close collaboration between the expertise from dynamical complex network analysis and modeling [Group of Nonlinear Dynamics Humboldt-Universität zu Beriin], biological psychology [Department of Psychology, Humboldt-Universität zu Beriin], and signal- and informafion processing in the neurosciences [Department of Computer Science,Humboldt-Universität zu Berlin], we will pursue the following goals:(i) Obtain high-resolution ERP data from psychological experiments with well-defined cognitive processes that are relevant for reading and (as benchmark) face processing.(ii) Extract functional and effective (causal) connectivity in sensor and source spaces to monitor the successive cognifive subprocesses of the brain activity during cognifion.The analysis of these large-scale brain networks will be performed by advanced toolsbased on the concept of hierarchical complex networks to characterise the fime evolving connecfivity associated with different experimental condifions.(iii) Develop biophysical complex dynamical network models that can account for ERP-like activity, representing the acfivation of brain regions involved in the processes understudy. The coupling parameters of such a model will allow us to identify the causal connectivity in the processing streamIn order to improve the analysis and the modeling, we will: (a) obtain structural MRT of the participants to be used in accurate source localisation of the brain activity, and (b) esfimate structural networks of the brains of the participants, using diffusion imaging such us DTI and/or HARDI or DSI and sufficient tractography methods to integrate the structural connectivity information into the network modeling.