Neural synchronization, brain networks, and cognitive processes

可以说，认知神经科学的主要任务是了解大脑如何实现认知过程，如感知、注意、记忆、决策和意识。我在这个提案中的目标是通过解决人类大脑区域网络中涉及实现认知过程的控制和信息处理的时间动态，在这个普遍问题上取得进展。我感兴趣的是确定这些网络的一般属性，这是至关重要的，同时也要确定它们的细节，这将有助于理解特定的过程，并在脑损伤康复等应用中发挥作用。我将通过研究脑电图（64通道）和脑磁图（151通道）的脑源之间的同步和格兰杰因果关系（时间序列的可预测性）以及脑源内部的同步来实现这一点，这些脑电图（64通道）和脑磁图（151通道）记录了许多认知过程，既可以通过进行实验来调用这些过程，也可以通过在理论上模拟大脑区域之间的交流来实现。脑电和脑电记录信号的脑源定位是一个难题，将采用几种互补的方法，包括波束形成（空间滤波）和独立分量分析（一种统计盲源分离技术），在一定程度上由类似范例的功能磁共振研究指导。测量同步的各种方法，包括小波变换和希尔伯特变换/分析信号分析，将被用来获得不同频段神经源活动的相位和幅度的时频肖像。此外，相位、幅度、同步和格兰杰因果关系的交叉频率分析将由定制软件完成。最后，在与通信工程师的合作下，我将在相关流程的各个阶段对这些区域之间的通信进行建模。同样，我们的目标是建立一个尽可能完整的大脑区域网络的时间动态图，这些区域参与了规范认知过程的实现。

Arguably the main task of cognitive neuroscience is to understand how the brain implements cognitive processes such as perception, attention, memory, decision making, and consciousness. My goal in this proposal is to make progress on this general problem by addressing the temporal dynamics of control and information processing within networks of human brain areas that are involved in implementing cognitive processes. I am interested both in specifying the general properties of these networks, which is of fundamental importance, as well as in actually specifying their details, which will be useful in understanding specific processes and in applications such as rehabilitation from brain injuries. I will do this by studying synchronization and Granger causality (predictability across time series) between, and also synchronization within, brain sources of EEG (64-channel) and MEG (151-channel) scalp-recorded signals for a number of cognitive processes, both empirically by conducting experiments that invoke those processes and theoretically by modeling communication between brain areas. Localization of brain sources of EEG- and MEG-recorded signals is a difficult problem and several complementary approaches will be taken, including beamforming (spatial filtering) and independent component analysis (a statistical blind source separation technique), guided to some extent by fMRI studies of similar paradigms. Various approaches to measuring synchronization, including wavelet and Hilbert transform/analytic signal analysis, will be used to obtain time-frequency portraits of phase and amplitude of the activity of the neural sources in various frequency bands. In addition, cross-frequency analysis of phase, amplitude, synchronization, and Granger causality will be accomplished by custom software. Finally, in collaboration with a communications engineer I will model the communications between these areas during various phases of the process in question. Again, the goal is to establish as complete a picture as possible of the temporal dynamics of the networks of brain areas that are involved in implementing canonical cognitive processes.