Quantifying Cortical Maturational Aspects During Different Vigilance States in Preterm Infants by Advanced EEG Analysis

Preterm birth is a significant risk factor for atypical neurodevelopment, yet early electrophysiological markers of brain maturation are still lacking. Non-invasive electroencephalographic (EEG) monitoring of cortical maturation in these patients holds promise as a tool for neurodevelopmental prediction. However, its clinical application is limited by technical challenges in maintaining stable, long-term electrode placement on very small neonate scalps and by the highly specialised, multi-level expertise required to care for these fragile patients. Using video-polysomnographic EEG recordings in very low birth weight (VLBW, < 1500 g) preterm infants, we characterised large-scale neuronal dynamics during distinct vigilance states and tested whether they could serve as indicators of early cortical maturation. We analysed EEG recordings obtained at 33.9 ± 1.4 weeks postmenstrual age (PMA), during active sleep (AS), sleep onset active sleep (SOAS), quiet sleep (QS), and quiet wakefulness (QW). For each vigilance state, we assessed large-scale neuronal dynamics in terms of phase synchronisation, neuronal bistability, and local phase-amplitude coupling (PAC), both globally and separately for anterior and posterior regions, and correlated them with PMA. We found that phase synchronisation peaked in the δ band during QS and in the θ band during more active states (QW, SOAS, AS). δ-band bistability was lower in posterior regions across all states, while δ-PAC was lower posteriorly during sleep but reversed during wakefulness. Also, bistability and PAC decreased with advancing PMA. These findings suggest that vigilance-state-dependent neuronal dynamics capture aspects of early cortical maturation—even with low-density EEG cap—offering novel candidate biomarkers to monitor neurodevelopment in infants born preterm.