Dynamic reorganization of intrinsic functional networks in the mouse brain

Functional connectivity (FC) derived from resting-state functional magnetic resonance imaging (rs-fMRI) allows for the integrative study of neuronal processes at a macroscopic level. The majority of studies to date have assumed stationary interactions between brain regions, without considering the dynamic aspects of network organization. Only recently has the latter received increased attention, predominantly in human studies. Applying dynamic FC (dFC) analysis to mice is attractive given the relative simplicity of the mouse brain and the possibility to explore mechanisms underlying network dynamics using pharmacological, environmental or genetic interventions. Therefore, we have evaluated the feasibility and research potential of mouse dFC using the interventions of social stress or anesthesia duration as two case-study examples. By combining a sliding-window correlation approach with dictionary learning, several dynamic functional states (dFS) with a complex organization were identified, exhibiting highly dynamic inter- and intra-modular interactions. Each dFS displayed a high degree of reproducibility upon changes in analytical parameters and across datasets. They fluctuated at different degrees as a function of anesthetic depth, and were sensitive indicators of pathology as shown for the chronic psychosocial stress mouse model of depression. Dynamic functional states are proposed to make a major contribution to information integration and processing in the healthy and diseased brain.

Traces Of Human Functional Activity: Moment-To-Moment Fluctuations In Fmri Data

Luca Dodero, Djalel Eddine Meskaldji, Dimitri Nestor Alice Van De Ville

Dynamic functional connectivity (dFC) measured by functional magnetic resonance imaging (fMRI) shows evidence of large-scale networks with highly dynamic (re) configurations. We propose a novel approach to extract traces of human brain function by the construction of a trajectory in a meaningful low-dimensional space. This allows studying dFC in more detail and identify possible meaningful brain states from the moment-to-moment fluctuations of the brain signals during resting state or naturalistic conditions such as passive movie watching. Specifically, we explored dynamic organization of sub-networks derived from the time-dependent graph Laplacian in combination with Riemannian manifold distance to measure dissimilarity over time of dFC and to subsequently build the trajectory of brain activity. As a proof-of-principle, we show results for an fMRI dataset containing both rest and movie epochs in 15 healthy participants. The movie condition varied (i.e., fearful, joyful, and neutral movie excerpts) and clearly influenced the subsequent resting-state period in terms of FC brain state.

IEEE2016

Development and application of dynamic functional connectivity methods to elucidate the neural underpinnings of human behaviour with functional magnetic resonance imaging

Thomas William Arthur Bolton

Human behaviour is exquisitely complex, because it is staggeringly multi-facetted and subtly varies across individuals. Characterising its yet incompletely understood underlying biological mechanisms has profound clinical implications. Here, our goal was thus to study inter-individual behavioural variability. To accomplish this feat, we monitored the dynamics of brain activity through functional magnetic resonance imaging. The acquired signals only indirectly reflect neuronal mechanisms, and can be modulated by several confounding factors. A notorious one is head motion of the scanned volunteers. As an initial step, we addressed to what extent such movement would relate to inter-individual behavioural variability. We revealed strong, yet overlooked ties between a broad range of anthropometric, cognitive, sensory and clinical measures, and spatio-temporal head motion specificities. Next, as behaviour is an intrinsically dynamic process, we scrutinised the literature to pinpoint tailored time-resolved approaches for the analysis of brain activity. In a first application, we opted to track the dynamic reconfigurations of functional interactions (or dynamic functional connectivity) between brain regions over time in the context of an audiovisual stimulation paradigm. To mitigate the deleterious influences of head motion, computations were performed across subjects (a process known as inter-subject functional correlation). We discovered that some cues from the movie would be perceived homogeneously by a healthy population, while others would not. In addition, typically developing individuals exhibited more homogeneous brain activity than subjects diagnosed with autism spectrum disorder. Furthermore, the latter differed in their individual response to the movie as a function of the extent and balance of their socio-communicative and repetitive behaviour symptoms. In a second step, we attempted to link behaviour and brain activity in the absence of an overt task. We decomposed the interactions of a dorsolateral prefrontal region of interest with the rest of the brain into a set of co-activation patterns, and observed that their expression in the resting-state could predict attentional performance. As the spatio-temporal complexity of functional brain dynamics transpired from this analysis, we then set to develop a novel sparse coupled hidden Markov model framework that enables to quantify the dynamic interplays that take place between brain networks. To try and minimise spurious effects driven by head motion, we also considered an alternative methodological avenue, called graph signal processing, which enables to inject knowledge regarding brain structure (known to constrain neuronal activity) into the analyses. By studying functional brain dynamics from this viewpoint, we revealed an extensive array of links to behaviour. Furthermore, we also extended the originally static approach to the dynamic setting, and by this mean, further broadened the set of unraveled brain/behaviour relationships. Finally, we looked back at all the methods that we applied in theoretical terms and in their ability to predict inter-individual differences in behaviour. We concluded that they render complementary aspects of functional brain dynamics, and that in future years, the neuronal significance of brain/behaviour relationships should be further strengthened by pursuing more extended multimodal analyses.

EPFL2020

Development and application of dynamic functional connectivity methods to elucidate the neural underpinnings of human behaviour with functional magnetic resonance imaging

Thomas William Arthur Bolton

EPFL2020