Development of brain state dynamics involved in working memory

1220 

Abstract Submission 

ying he¹

¹McGill University, Montreal, Quebec

ying he  
McGill University
Montreal, Quebec

The human brain is an open complex system that is dynamically organized to support cognitive and behavioral flexibility, as well as rapid adaptation to ever-changing environmental or task demands. A wealth of recent neuroimaging studies in adults has demonstrated that the dynamical organization of functional brain networks is essential to support the performance of various cognitive and affective functions. Alterations in the dynamical organization of large-scale functional brain networks are linked to cognitive impairments in a variety of psychiatric and neurodevelopmental disorders. As childhood is a critical period during which the human brain function and structure undergo protracted development with prominent changes in higher-order cognitive functions such as working memory, it is thus pivotal to understand how the dynamical organization of functional brain networks contributes to children's cognitive development and the architecture of structural connectivity underlying brain dynamics.

To address above open questions, we set up a developmental functional Magnetic Resonance Imaging (fMRI) study with advanced analytic approaches and Diffusion Tensor Imaging (DTI) study in 69 children (aged 7–12-year-old) and 51 adults (aged 19–24-year-old) to investigate developmental differences in dynamical brain states of the frontal-parietal network (FPN) and default mode network (DMN) during working memory task and white matter structural connectivity. Observed fMRI data are modeled by a novel Bayesian switching dynamical systems (BSDS) approach. The BSDS model was applied to the time courses which could provide the temporal evolution of the states, the occupancy rate and mean lifetime of states, the transition probability matrix and mean and covariance of states. For each subject, we also calculated the mean fractional anisotropy (FA) values across all voxels in the tract between each two ROIs to weighted the white matter connectivity matrix. And then, the structure-function coupling was measured as Spearman rank correlation between nonzero element of structural and functional connectivity profiles.

We identified five brain states with rapid transitions, characterized by dynamic configurations among FPN and DMN nodes with active and inactive engagement in different task demands. Compared with adults, children exhibited less brain states with highest activity in FPN nodes dominant to high demand, and its occupancy rate increased with age. Children preferred to attain inactive brain states with low activity in both FPN and DMN nodes. Moreover, children exhibited lower transition probability from low-to-high demand states and such transition was positively related with working memory performance. Notably, higher transition probability from low-to-high demand states was associated with stronger structural connectivity across FPN and DMN, but weaker structure-function coupling of these two networks.

In conclusion, our study demonstrates immature dynamical organization of the FPN and DMN nodes during WM, characterized by less occupancy rate and more transitions to inactivated state but weaker transitions among brain states dominant to high task demand condition. And effective brain dynamics is related with stronger structural connectivity but weaker structure-function coupling among FPN and DMN. Our findings extend our current understanding of how the FPN and DMN nodes are dynamically organized into a set of nuanced brain sates to support moment-to-moment information updating during WM and its links to structural connectivity.

Executive Function, Cognitive Control and Decision Making

Working Memory ²

Early life, Adolescence, Aging ¹

Connectivity (eg. functional, effective, structural)

Other - brain states, working memory, development, structural connectivity