The role of structural heterogeneity in shaping whole-brain dynamics across scales

The brain must rapidly and dynamically recruit the appropriate brain regions to respond to environmental changes. An extensive body of literature views the ability to spontaneously explore a rich repertoire of activity patterns, even during rest, as a correlate of the capacity to respond rapidly to internal fluctuations and external stimuli. Brain flexibility, defined as the number of unique patterns of activity explored, has been used to quantify the richness of resting-state dynamics. Beyond its theoretical relevance, brain flexibility successfully indexes pathological conditions such as Parkinsons disease, multiple sclerosis, and amyotrophic lateral sclerosis. However, this approach is inherently global and provides no information about the contribution of each region to the flexibility of brain dynamics. Given the functional and structural heterogeneity of brain regions, we expect their roles in sustaining flexibility to be heterogeneous and related to regional structural properties. To investigate these hypotheses, we present the flexibility gradient (FG), which quantifies regional contributions to flexibility. To quantify FG across the brain, we employed three datasets spanning different recording techniques and a range of spatial and temporal scales. We used 47 MEG and 11 EEG recordings to access fast dynamics, while to investigate slower dynamics, we leveraged 77 fMRI recordings along with the corresponding tractographies. We found that FG is not homogeneously distributed, with associative regions contributing the most. The same gradient is evident within each hemisphere independently, and the structural properties of brain regions correlate with the FG. Our findings are stable across modalities and time scales.