How do spatially distinct frequency specific MEG networks emerge from one underlying structural connectome? The role of the structural eigenmodes

Abstract

Functional networks obtained from magnetoencephalography (MEG) from different frequency bands show distinct spatial patterns. It remains to be elucidated how distinct spatial patterns in MEG networks emerge given a single underlying structural network. Recent work has suggested that the eigenmodes of the structural network might serve as a basis set for functional network patterns in the case of functional MRI. Here, we take this notion further in the context of frequency band specific MEG networks. We show that a selected set of eigenmodes of the structural network can predict different frequency band specific networks in the resting state, ranging from delta (1-4Hz) to the high gamma band (40-70Hz). These predictions outperform predictions based from surrogate data, suggesting a genuine relationship between eigenmodes of the structural network and frequency specific MEG networks. We then show that the relevant set of eigenmodes can be excited in a network of neural mass models using linear stability analysis only by including delays. Excitation of an eigenmode in this context refers to a dynamic instability of a network steady state to a spatial pattern with a corresponding coherent temporal oscillation. Simulations verify the results from linear stability analysis and suggest that theta, alpha and beta band networks emerge very near to the bifurcation. The delta and gamma bands in the resting state emerges further away from the bifurcation. These results show for the first time how delayed interactions can excite the relevant set of eigenmodes that give rise to frequency specific functional connectivity patterns.

Optimization of graph construction can significantly increase the power of structural brain network studies

“…One of the main goals of brain network analyses is to understand the structural underpinnings of the functional organization of the human brain. Functional networks generated from magnetoencephalography or electroencephalography data have revealed that the strength of functional connections between brain areas depends on the frequency of the signals (Brookes et al, 2011;Deco et al, 2017;Messaritaki et al, 2017;Tewarie et al, 2019). Additionally, they have shown that there may be more than one mechanism of coupling between brain areas, such as phasephase, phase-frequency, phase-amplitude and amplitude-amplitude (Hyafil et al, 2015;Dimitriadis et al, 2015;Dimitriadis and Salis, 2017;Dimitriadis, 2018;.…”

Using structural connectivity to augment community structure in EEG functional connectivity

“…Understanding how flexible macroscopic functional architecture arises from the stable scaffold of anatomical white matter fiber tracts in the brain is a major field of inquiry in neuroscience. This has been investigated via the relationship between structural and functional connectivity (SC and FC, respectively) between brain regions in fMRI and MEG (Abdelnour, Dayan, Devinsky, Thesen, & Raj, 2018;Atasoy, Donnelly, & Pearson, 2016;Cabral et al, 2014;Damoiseaux & Greicius, 2009;Deco et al, 2013;Glomb, Ponce-Alvarez, Gilson, Ritter, & Deco, 2017;Goñi et al, 2014;Hagmann et al, 2008;Honey et al, 2009;Meier et al, 2016;Tewarie et al, 2019Tewarie et al, , 2014Vincent et al, 2007) . Concurring findings from these studies show that FC between regions of interest (ROIs)/sources located in the gray matter is in part shaped by anatomical connections of the SC, such that the strength of SC (fiber count, density) is predictive to some degree of the strength of FC (correlation, coherence, etc.)…”

Multichannel Time–Frequency Complexity Measures for the Analysis of Age-Related Changes in Neuromagnetic Resting-State Activity

“…This result evidences the limitations of classical power spectral analysis, while possing interpretations issues: three different results related to three different frequency bands that must be discussed. The interpretation of different band-specific results emerging from a single structural network is still an ongoing debate [42]. On the contrary, the analysis proposed in this work does not rely on classical frequency bands, offering a single result.…”