Network analysis of whole-brain fMRI dynamics: A new framework based on dynamic communicability

Abstract

Neuroimaging techniques such as MRI have been widely used to explore the associations between brain areas. Structural connectivity (SC) captures the anatomical pathways across the brain and functional connectivity (FC) measures the correlation between the activity of brain regions. These connectivity measures have been much studied using network theory in order to uncover the distributed organization of brain structures, in particular FC for task-specific brain communication. However, the application of network theory to study FC matrices is often "static" despite the dynamic nature of time series obtained from fMRI. The present study aims to overcome this limitation by introducing a network-oriented analysis applied to whole-brain effective connectivity (EC) useful to interpret the brain dynamics. Technically, we tune a multivariate Ornstein-Uhlenbeck (MOU) process to reproduce the statistics of the whole-brain resting-state fMRI signals, which provides estimates for MOU-EC as well as input properties (similar to local excitabilities). The network analysis is then based on the Green function (or network impulse response) that describes the interactions between nodes across time for the estimated dynamics. This model-based approach provides time-dependent graph-like descriptor, named communicability, that characterize the roles that either nodes or connections play in the propagation of activity within the network. They can be used at both global and local levels, and also enables the comparison of estimates from real data with surrogates (e.g. random network or ring lattice). In contrast to classical graph approaches to study SC or FC, our framework stresses the importance of taking the temporal aspect of fMRI signals into account. Our results show a merging of functional communities over time (in which input properties play a role), moving from segregated to global integration of the network activity. Our formalism sets a solid ground for the analysis and interpretation of fMRI data, including task-evoked activity. 15 to evaluate global effects in recurrently-connected networks [9]. It has then been used to derive measures that describe the roles for nodes in networks [10] and to define a version for centrality in graphs [11,12]. In the context of neuroimaging, graph communicability has been applied to evaluate the contribution of SC topology in generating FC [13]. However, graph theory is often applied in an off-the-shelf manner and this type of approach is limited 20 for explaining the time-series nature of the fMRI measurements.In this context the present study aims to describe the fMRI-related functional associations between the nodes in the brain network, also known as regions of interest (ROIs), while properly taking time into account. We follow recent works that employed dynamic models of the brain activity to link SC and FC [14]. A great variety of network designs has been explored 25 to combine experimental data in various levels of detail [15,16,17,18,19, 20]. These dynamic models typically involve a ...

Brain Connectivity Studies on Structure‐Function Relationships: A Short Survey with an Emphasis on Machine Learning

“…In an effort to explain brain dynamics, Gilson et al [ 63 ] considered a multivariate Ornstein–Uhlenbeck model (An Ornstein–Uhlenbeck process is a stochastic stationary Gauss–Markov process, which solves the Langevin equation.) [ 64 ] on a DGM to estimate statistics characterizing effective connectivities (ECs) in RSNs.…”

Static and dynamic functional connectivity supports the configuration of brain networks associated with creative cognition

“…Detecting various communities and dividing the network into sub-networks by maximizing the modularity function is an important feature of network analysis 63 , 76 , 77 . To understand the variation in static network, dynamic networks, creative task and rest we used a popular and fast iterative algorithm called the fast greedy community detection algorithm from previous fMRI studies 78 – 81 . This algorithm uses a hierarchical and bottom-up approach to optimize the modularity function in detecting communities within complex brain networks 82 – 84 .…”