Mechanistic insights into cortical network entrainment by tACS and tDCS: A computational modeling approach

Non-invasive brain stimulation techniques such as transcranial alternating current stimulation (tACS) and transcranial direct current stimulation (tDCS) are increasingly explored for modulating large-scale brain oscillations associated with cognition, sleep, and neuropsychiatric disorders. However, the network-level mechanisms underlying their differential effects on synchronization and excitatory-inhibitory (E/I) dynamics remain poorly understood. This study investigates how tDCS, tACS, and algorithmically guided tACS (ALGO-tACS) influence phase synchronization, intrinsic frequency modulation, phase-amplitude coupling (PAC), and E/I balance in a large-scale cortical network model. A biologically plausible network of 1000 Izhikevich neurons (800 excitatory, 200 inhibitory) was simulated to generate intrinsic delta (3 Hz) oscillations. Stimulation protocols included continuous depolarizing and hyperpolarizing tACS, tDCS, and adaptive ALGO-tACS, matched to intrinsic network phase and frequency. Depolarizing tACS enhanced intrinsic delta-band power (~3-4 Hz) and elevated PLV (~0.99), indicating effective phase entrainment. In contrast, hyperpolarizing tACS disrupted synchrony and reduced oscillatory power. tDCS enhanced PAC (MI: 0.0180), indicating stronger delta-gamma coupling, but reduced PLV and overall coherence. ALGO-tACS combined near-perfect PLV (~0.99), strong delta coherence, and the highest PAC (MI: 0.0183), reflecting superior adaptive modulation of network dynamics. Cross-coherence analysis revealed that tACS and ALGO-tACS preserved low-frequency coherence across excitatory-inhibitory populations, while tDCS disrupted higher-frequency synchrony. These findings demonstrate that phase- and frequency-matched stimulation, especially ALGO-tACS, optimizes network synchronization and cross-frequency coupling, preserving E/I balance. This mechanistic framework supports the development of personalized, closed-loop neuromodulation strategies targeting sleep, cognition, and neuropsychiatric disorders. • Developed a biologically inspired cortical network model (800 PY, 200 IN) exhibiting intrinsic delta oscillations. • Compared tDCS, tACS, and algorithm-enhanced tACS (ALGO-tACS) in modulating synchronization and PAC. • ALGO-tACS achieved near-perfect PLV (~0.99) and highest PAC (MI=0.0183), combining tACS and tDCS benefits. • Cross-coherence analysis revealed tACS and ALGO-tACS preserved delta coherence, while tDCS disrupted higher-frequency synchrony. • Results suggest adaptive stimulation protocols optimize network-level dynamics, offering potential for targeted neuromodulation in sleep and cognition.