Applications of Community Detection Techniques to Brain Graphs: Algorithmic Considerations and Implications for Neural Function

Abstract

The human brain can be represented as a graph in which neural units such as cells or small volumes of tissue are heterogeneously connected to one another through structural or functional links. Brain graphs are parsimonious representations of neural systems that have begun to offer fundamental insights into healthy human cognition, as well as its alteration in disease. A critical open question in network neuroscience lies in how neural units cluster into densely interconnected groups that can provide the coordinated activity that is characteristic of perception, action, and adaptive behaviors. Tools that have proven particularly useful for addressing this question are community detection approaches, which can identify communities or modules: groups of neural units that are densely interconnected with other units in their own group but sparsely interconnected with units in other groups. In this paper, we describe a common community detection algorithm known as modularity maximization, and we detail its applications to brain graphs constructed from neuroimaging data. We pay particular attention to important algorithmic considerations, especially in recent extensions of these techniques to graphs that evolve in time. After recounting a few fundamental insights that these techniques have provided into brain function, we highlight potential avenues of methodological advancements for future studies seeking to better characterize the patterns of coordinated activity in the brain that accompany human behavior. This tutorial provides a naive reader with an introduction to theoretical considerations pertinent to the generation of brain graphs, an understanding of modularity maximization for community detection, a resource of statistical measures that can be used to characterize community structure, and an appreciation of the usefulness of these approaches in uncovering behaviorally-relevant network dynamics in neuroimaging data.

Reconfigurations within resonating communities of brain regions following TMS reveal different scales of processing

“…These two parameters determine 204 the scale of the resulting graph, both structurally and temporally. As described in Garcia et al (2018), 205…”

“…These two parameters determine the scale of the resulting graph, 661 both structurally and temporally, and here, we sweep this parameter space to find the scale of the 662 data that is most unlike that expected in an appropriate random network null model. As described in 663 Garcia et al (2018), there are several heuristics we may use to determine the optimal parameter for 664 our dataset. We chose an unbiased "difference" heuristic because of the unique properties of this 665 stimulation dataset, which we explain below.…”

“…interacting networks, few studies have employed the rich set of tools of network science to understand 98 the propagation of stimulation (Bortoletto et al, 2015). Network science not only provides a 99 mathematical language to describe complex connectivity patterns resulting from stimulation, but 100 previous research has also proposed a variety of summary metrics in which to characterize local and 101 global connectivity in the brain (for review, see Garcia et al, 2018). In this study, we address this 102 existing gap in the literature and employ a method recently developed in network science to study the 103 interactions of local connectivity and global network dynamics.…”

Distributed Subnetworks of Depression Defined by Direct Intracranial Neurophysiology

“…To tackle inconsistent network sampling across individuals, we utilized a novel method called SuperEEG 47 that uses the correlational relationship of brain activity across electrodes to generate a whole-brain iEEG model for each subject. We next uncovered the inherent organization of the brain's correlational structure from the output of this model by examining network modularity, a graph theoretic concept that quantifies the organization of a system into constituent modules 49,50 , and has been used to reveal system disruptions in disease states [51][52][53][54][55][56][57] including MDD 29 . This approach revealed 6 functional networks in our data, which we refer to as "modules.…”

“…First, the population-level correlational model was reordered by the module assignment according to the modularity cost function. Network modules were identified at different levels of granularity by varying the tuning parameter, γ 29,49 . Increasing γ partitions the brain into increasing numbers of modules with a limit equal to the number of electrodes, as shown here for 3 values of γ (left).…”