Periodic oscillatory information flow across brain networks underlies cortical-wide neuronal avalanches

The brain consists of local and large-scale functional neural networks that must tightly communicate to allow information processing and integration. Neuronal avalanches are neural activities occurring at near-critical state of the brain, which enhances its capacity to carry and process information. However, the relationship between functional networks and neuronal avalanches remains largely unknown. Here, human magnetoencephalography data containing information of both cortical-wide avalanches and oscillatory phases of information flow were utilized to explore this relationship. We found that cortical-wide avalanches of specific frequency bands are strongly associated with information flow across anteroposterior cortical gradients, which connect important hubs within known functional networks. Also, long-range memories of local oscillatory activities at these network hubs strongly influence global avalanche activities. By computer modelling, we demonstrated that periodic, but not aperiodic, oscillations of specific frequencies generated at network hubs along a cortical gradient could explain the occurrence of avalanches. This study represents the first attempt to integrate previously disparate explorations of functional networks and neuronal avalanches under a unified framework. Consequently, it provides mechanistic insights into how oscillatory information flow underlies neuronal avalanches to facilitate cortical-wide communication, which also implicates the significance of different frequencies of brain activities in health and neuropsychiatric conditions.