Dynamic network reorganization underlying neuroplasticity: the deficits-severity-related language network dynamics in patients with left hemispheric gliomas involving language network.

Brain network dynamics not only endow the brain with flexible coordination for various cognitive processes but also with a huge potential of neuroplasticity for development, skill learning, and after cerebral injury. Diffusive and progressive glioma infiltration triggers the neuroplasticity for functional compensation, which is an outstanding pathophysiological model for the investigation of network reorganization underlying neuroplasticity. In this study, we employed dynamic conditional correlation to construct framewise language networks and investigated dynamic reorganizations in 83 patients with left hemispheric gliomas involving language networks (40 patients without aphasia and 43 patients with aphasia). We found that, in healthy controls (HCs) and patients, the language network dynamics in resting state clustered into 4 temporal-reoccurring states. Language deficits-severity-dependent topological abnormalities of dFCs were observed. Compared with HCs, suboptimal language network dynamics were observed for those patients without aphasia, while more severe network disruptions were observed for those patients with aphasia. Machine learning-based dFC-linguistics prediction analyses showed that dFCs of the 4 states significantly predicted individual patients' language scores. These findings shed light on our understanding of metaplasticity in glioma. Glioma-induced language network reorganizations were investigated under a dynamic "meta-networking" (network of networks) framework. In healthy controls and patients with glioma, the framewise language network dynamics in resting-state robustly clustered into 4 temporal-reoccurring states. The spatial but not temporal language deficits-severity-dependent abnormalities of dFCs were observed in patients with left hemispheric gliomas involving language network. Language network dynamics significantly predicted individual patients' language scores.