Signaling in brain networks unfolds over multiple topological scales. Areas may exchange information over local circuits, encompassing direct neighbours and areas with similar functions, or over global circuits, encompassing distant neighbours with dissimilar functions. Here we study how the organization of cortico-cortical networks mediate localized and global communication by parametrically tuning the range at which signals are transmitted on the white matter connectome. We show that brain regions vary in their preferred communication scale. By investigating the propensity for brain areas to communicate with their neighbors across multiple scales, we naturally reveal their functional diversity: unimodal regions show preference for local communication and multimodal regions show preferences for global communication. We show that these preferences manifest as region- and scale-specific structure-function coupling. Namely, the functional connectivity of unimodal regions emerges from monosynaptic communication in small-scale circuits, while the functional connectivity of transmodal regions emerges from polysynaptic communication in large-scale circuits. Altogether, the present findings reveal that communication preferences are highly heterogeneous across the cortex, shaping regional differences in structure-function coupling.

大脑网络中的信号在多个拓扑尺度上展开。区域可以通过局部电路交换信息，包括直接邻居和具有相似功能的区域，或者通过全局电路交换信息，包括具有不同功能的远邻。在这里，我们研究了皮质 - 皮质网络的组织如何通过参数调整信号在白质连接组上传输的范围来介导局部和全局通信。我们表明，大脑区域的首选交流规模各不相同。通过调查大脑区域在多个尺度上与其邻居交流的倾向，我们自然地揭示了它们的功能多样性：单峰区域显示出对局部交流的偏好，而多峰区域则显示出对全局交流的偏好。我们表明，这些偏好表现为区域和尺度特定的结构-功能耦合。即，单峰区域的功能连接来自小规模电路中的单突触通信，而跨模式区域的功能连接来自大规模电路中的多突触通信。总之，目前的研究结果表明，整个皮层的交流偏好是高度异质的，从而形成了结构-功能耦合的区域差异。