Department of Physiology, Semmelweis University, 37-43 Tűzoltó Street, 1094 Budapest, Hungary. Institute of Clinical Experimental Research, Semmelweis University, 37-43 Tűzoltó Street, Budapest 1094, Hungary.
Physiol Meas. 2018 Feb 28;39(2):024003. doi: 10.1088/1361-6579/aaa916.
Brain function is organized as a network of functional connections between different neuronal populations with connection strengths dynamically changing in time and space. Studies investigating functional connectivity (FC) usually follow a static approach when describing FC by considering the connectivity strengths constant, however a dynamic approach seems more reasonable, as this way the spatio-temporal dynamics of the underlying system can also be captured.
The scale-free, i.e. fractal nature of neural dynamics is an inherent property of the nervous system. The aim of this study was to determine if dynamic functional connectivity (DFC) in the prefrontal cortex shows not only scale-free but indeed multifractal dynamics.
Functional near-infrared spectroscopy (fNIRS) was used to monitor resting-state brain activity in young healthy volunteers. Sliding window correlation (SWC) analysis and graph theory approach were utilized to capture the functional connection networks for every time point, whose topology was subsequently characterized with three network metrics-Density, Clustering Coefficient and Efficiency-each capturing a different aspect of the given network. The temporal structuring of the obtained network metric time series was then described by multifractal time series analysis.
We found the DFC in the prefrontal cortex fluctuating according to scale-free, specifically multifractal dynamics. Moreover, different topological properties of the network showed different multifractal characteristics. All the results were reproducible in all window sizes used in the SWC analysis, however we found that the actual values of the given multifractal properties depended significantly on the window size.
Our results may well be another indication of a self-organized critical state underlying resting-state brain activity. The proposed analysis of functional brain dynamics can also open new perspectives for future clinical applications.
大脑功能组织为不同神经元群体之间的功能连接网络，连接强度随时间和空间动态变化。研究功能连接（FC）的通常采用静态方法来描述 FC，认为连接强度是恒定的，但动态方法似乎更为合理，因为这样可以捕捉到基础系统的时空动态。
神经动力学的无标度，即分形性质是神经系统的固有属性。本研究旨在确定前额叶皮层的动态功能连接（DFC）是否不仅具有无标度特性，而且实际上具有多重分形动力学特性。
使用功能近红外光谱（fNIRS）监测年轻健康志愿者的静息态大脑活动。滑动窗口相关（SWC）分析和图论方法用于捕获每个时间点的功能连接网络，其拓扑结构随后用三个网络度量来描述-密度、聚类系数和效率-每个度量都捕捉到给定网络的不同方面。获得的网络度量时间序列的时间结构随后通过多重分形时间序列分析来描述。
我们发现前额叶皮层的 DFC 根据无标度，特别是多重分形动力学而波动。此外，网络的不同拓扑性质显示出不同的多重分形特征。所有结果在 SWC 分析中使用的所有窗口大小上都具有可重复性，但我们发现，给定多重分形性质的实际值取决于窗口大小。
我们的结果可能是静息态大脑活动潜在的自组织临界状态的另一个迹象。所提出的功能大脑动力学分析也为未来的临床应用开辟了新的视角。
