Spectral dynamic causal modeling of effective connectivity across multiple brain networks in pilot trainees

Pilots operate in complex, time-varying environments, relying on specific causal brain-network dynamics to support their cognitive skills. Resting-state functional magnetic resonance imaging and structural imaging were collected from healthy male flight cadets (n = 39) and healthy male controls (n = 37), alongside a card-sorting measure of executive function and a visual change-detection task. Spectral dynamic causal modeling with hierarchical Bayesian inference was used to quantify effective connectivity within and between large-scale networks. Compared with controls, flight cadets exhibited significant differences in within-network and between-network effective connectivity spanning the default mode, sensorimotor, ventral attention, dorsal attention, and visual networks. Connectivity-behavior relationships also differed between groups, including a flight-cadet-specific positive association within the ventral attention network. These findings link long-term flight training with distinctive patterns of brain-network causal interactions related to executive performance, and offer a potential biological perspective for understanding neuroadaptation during early-stage flight training and for informing training optimization.