Neurodynamics of prefrontal areas in volitional contexts: a comparative study based on computational modelling and EEG/ERP data

Volitional action arises from the interaction between internal intentions and external stimuli, yet the neural dynamics distinguishing self-initiated from externally triggered actions remain unclear. Here, we combine human EEG/ERP data with simulations from a phenomenological neurocomputational model of prefrontal control to examine the neurodynamic principles underlying different types of volitional behaviors. Using an extended version of our previously developed model, we generated neural activity patterns corresponding to self-initiated and externally triggered actions by manipulating the balance between endogenous and exogenous inputs to lateral prefrontal subregions. We then qualitatively compared these simulated dynamics with empirical EEG data from a perceptual decision-making task involving voluntary and instructed skip actions. Across both datasets, self-initiated actions showed a gradual buildup of activity, marked reductions in cross-trial variability, sharper state transitions, and beta-gamma frequency shifts; in contrast, externally triggered actions exhibited minimal variability reduction, weaker transitions, and largely stable frequency content. These convergent results suggest that self-initiated actions are supported by distinct preparatory dynamics, characterized by internal competition and stabilization of neural states, whereas externally triggered actions rely primarily on externally driven input. Together, these findings identify candidate neurodynamic signatures of intention formation and highlight how simplified computational models can reproduce and help explain key features of volitional control observed in human electrophysiology.