Structural variability in the human brain reflects fine-grained functional architecture at the population level

Any further question? Gwenaëlle Douaud (Oxford FMRIB Centre, Wellcome Centre for Integrative Neuroimaging)

Human brain structure topography is thought to be related in part to functional specialisation. However, the extent of such relationships is unclear. Here, using a data-driven, multi-modal approach for studying brain structure across the lifespan (n=484, 260 females), we demonstrate that numerous structural networks, covering the entire brain, follow a functionally-meaningful architecture. These grey matter networks emerge from the co-variation of grey matter volume and cortical area at the population level. We further reveal fine-grained anatomical signatures of functional connectivity. For example, within the cerebellum, a structural separation emerges between lobules that are functionally connected to distinct, mainly sensorimotor, cognitive and limbic regions of the cerebral cortex and subcortex. Structural modes of variation also replicate the fine-grained functional architecture seen in 8 well-defined visual areas in both task and resting-state fMRI. Furthermore, our study shows a structural distinction corresponding to the established segregation between anterior and posterior default-mode networks. These fine-grained grey matter networks further cluster together to form functionally-meaningful larger-scale organisation. In particular, we identify a structural architecture bringing together the functional posterior default-mode network and its anti-correlated counterpart. In summary, our results demonstrate that the relationship between structural and functional connectivity is fine-grained, widespread across the entire brain, and driven by co-variation in cortical area, i.e. likely differences in shape, depth or number of foldings. These results suggest that neurotrophic events occur during development to dictate that the size and folding pattern of distant, functionally-connected brain regions should vary together across subjects.

Smith S, Duff E, Groves A, Nichols TE, Jbabdi S, Westlye LT, Tamnes CK, Engvig A, Walhovd KB, Fjell AM, Johansen-Berg H, Douaud G.
Structural variability in the human brain reflects fine-grained functional architecture at the population level
J Neurosci, 2019 39(31) 6136-6149.

Results and Script

• Results report with all 70 components from the FLICA decomposition of cortical thickness, cortical area and GM volume (VBM)
  across all 484 healthy subjects, including post-hoc associations with age, sex and intracranial volume (NB: index starts at 0):
  • Report
• Script with MATLAB and FSL commands for running FLICA* and estimating the variance explained by each component
  (see Table 1 of our paper; NB: index starts at 1):
  • Script
• Result maps of all 70 components: cortical thickness, cortical area and GM volume (VBM) across all 484 healthy subjects:
  • Cortical thickness decomposition in volume (approximation)
  • FreeSurfer cortical thickness decomposition, L
  • FreeSurfer cortical thickness decomposition, R
  • Cortical area decomposition in volume (approximation)
  • FreeSurfer cortical area decomposition, L
  • FreeSurfer cortical area decomposition, R
  • FSL-VBM GM volume
• Parcellation map of all 36 components:
  • Hard parcellation (winner-take-all) created using the absolute z maps for the 36 structural GMNs used in the hierarchical clustering analysis

  * The FLICA code itself can be found here