Networks of myelin covariance

Abstract

Networks of anatomical covariance have been widely used to study connectivity patterns in both normal and pathological brains based on the concurrent changes of morphometric measures (i.e., cortical thickness) between brain structures across subjects (Evans, 2013). However, the existence of networks of microstructural changes within brain tissue has been largely unexplored so far. In this article, we studied in vivo the concurrent myelination processes among brain anatomical structures that gathered together emerge to form nonrandom networks. We name these “networks of myelin covariance” (Myelin‐Nets). The Myelin‐Nets were built from quantitative Magnetization Transfer data—an in‐vivo magnetic resonance imaging (MRI) marker of myelin content. The synchronicity of the variations in myelin content between anatomical regions was measured by computing the Pearson's correlation coefficient. We were especially interested in elucidating the effect of age on the topological organization of the Myelin‐Nets. We therefore selected two age groups: Young‐Age (20–31 years old) and Old‐Age (60–71 years old) and a pool of participants from 48 to 87 years old for a Myelin‐Nets aging trajectory study. We found that the topological organization of the Myelin‐Nets is strongly shaped by aging processes. The global myelin correlation strength, between homologous regions and locally in different brain lobes, showed a significant dependence on age. Interestingly, we also showed that the aging process modulates the resilience of the Myelin‐Nets to damage of principal network structures. In summary, this work sheds light on the organizational principles driving myelination and myelin degeneration in brain gray matter and how such patterns are modulated by aging.

The developing relations between networks of cortical myelin and neurophysiological connectivity

“…Our data support the use of T1-w/T2-w ratio to measure myelin covariance -our structural covariance networks demonstrated organization as well as high and symmetric intra-hemispheric connectivity in both children and adults, consistent with our previous study using MT (15). Myelin covariance is highest between the bilateral subcortical and limbic structures, both within and between hemispheres, and amongst within-hemispheric connections in the occipital and parietal lobes (particularly in the adults), as seen by Melie-Garcia et al (24).…”

“…Consistent with previous studies examining typical development from childhood to young adulthood, the age-related changes of both structural covariance (24)(25)(26)(27) and MEG functional connectivity (28)(29)(30) follow predominantly nonlinear, quadratic trajectories. The differences in functional connectivity with age was marked by increases in functional connectivity with age in the alpha and beta bands and decreases in the gamma bands, consistent with developmental changes reported in both phase-based (31) and amplitude (30, 32) connectivity measures.…”

“…To construct age-dependent networks, a sliding window approach was employed (24,25). The number of participants per window (window width) was held constant at 50, well above the minimum number of participants necessary for reliable correlations when calculating structural covariance (75).…”

White Matter Plasticity Keeps the Brain in Tune: Axons Conduct While Glia Wrap

“…Rather, it has been observed that distinct networks tend to covary with respect to myelin changes with age ( Evans, 2013 ). Interestingly, these network correlations in myelination changes were shown to potentially increase the resilience of covarying networks to damage, using simulations in which network hubs were deleted ( Melie-Garcia et al, 2018 ). Thus, the specific changes in connectivity that occur in the aging brain may confer improved resistance to increasingly probable network damage as individuals age.…”

The Inequality of Neural Destiny: Signatures of Life course Socioeconomic Conditions in Brain Myelination and Grey Matter Volume

“…) and myelin content(35)). Crucially, quantitative MRI maps are impervious to inter-site variability ( (49)).…”