Functional connectome of human cerebellum

Abstract

Background Neural connectome theory has been widely used in system neuroscience, and prompted our comprehension for the topological organizations of human cerebral cortex. However, how functional connectome is organized topologically in cerebellums remains unclear. Method Resting-state functional connectivity (rs-fcMRI) data were acquired from 1416 healthy adults in two independent samples. In Sample 1 (n = 976), both voxel-wise and node-wise topological properties for functional cerebellar connectome were estimated. Moreover, the network-based topological properties of cerebellum and cerebro-cerebellar topological mapping were investigated, respectively. Given the temporal natures in the neural population, a hidden Markov model (HMM) was further capitalized to uncover the dynamic pattern of cerebellar functional connectome. In order to test the robustness of our findings, we ran all of the analyses in an independent dataset (Sample 2; n = 440). Results We found that Crus I and II exhibited prominently high degree centrality (DC) for cerebellar functional connectome. Further, the cerebellar functional connectome and even the nested network-wise cerebellar connectome was found to be organized by small-world, modular and hierarchical manners significantly. Also, three intrinsic modules were found in cerebellar functional connectome, including attention/executive network, default mode network and task-positive network. In addition, the significant cerebro-cerebellar correlations for small-world organization and hierarchical architecture were found as well. By building cerebro-cerebellar topological mapping, both frontoparietal and subcortical networks were found to be overrepresented into cerebellums than cerebral cortex (3-fold). As for temporal natures, cerebellar functional connectome was observed to be highly flexible and modular, but showed high individual-specific variances in temporal dynamic pattern. Conclusion This study identified the topological architectures and temporal hierarchy of functional cerebellar connectome, and further demonstrated prominent functional cerebro-cerebellar couplings of small-world organization and hierarchical architectures.

Keywords: Cerebellum; Hidden Markov model; Neural connectome; Neural modules; Temporal structures.