{"paper_id":"314152c7-c3ab-4447-9e3a-5e22eb10f079","body_text":"Abstract\nTraveling waves of neural activity are fundamental to cortical function, yet their spatiotemporal dynamics and structural constraints remain less understood in human brain. Conventional analyses focus on network topology and spatial geometry of neural signals, with limited emphasis on wave dynamics at high spatiotemporal resolutions. Here, we applied a manifold-based optical flow algorithm to an open-access intracranial electroencephalography (iEEG) dataset from the single-pulse stimulation cortico-cortical evoked potentials (CCEPs) paradigm, enabling direct construction of high-temporal-resolution cortical neural activity velocity fields. Spatiotemporal mode decomposition revealed a hierarchical organization: large-scale plane waves (mediating global propagation) dominate, complemented by local patterns for local computation. Dominant plane wave directions correlate with underlying white matter fibre tract orientations, indicating anatomical guidance of cortical dynamics. Velocity field singularities (sources, sinks, spirals and saddles) are non-randomly distributed, associating with default mode network (DMN) hubs. We further identified a nonlinear relationship between cortical wave speed and age, with developmental phases paralleling white matter maturation and aging. Our findings link cortical wave dynamics to anatomy, functional networks, and neurodevelopment, establishing wave speed as a potential yet important biomarker for brain function and health.\nCompeting Interest Statement\nThe authors have declared no competing interest.","source_license":"CC-BY-4.0","license_restricted":false}