Electroencephalographic functional connectivity in extreme prematurity: a pilot study based on graph theory

Background: Connectivity studies based on functional magnetic resonance imaging (MRI) provided new insights in neonatal brain development but cannot be performed at bedside in the clinical setting. The electroencephalogram (EEG) connectivity has been less studied, particularly using the new approach based on graph theory. This study aimed to explore the functional EEG connectivity using graph theory analysis at an early post-conception age in extremely premature and late-preterm babies free of medical complications and overt brain damage. Methods: Sixteen neonates (8 extremely low gestational age (ELGA) and 8 late-preterm infants), both groups having performed multichannel EEG recordings at 35 weeks’ post-conception, were recruited in a single tertiary-level neonatal intensive care unit and well-baby nursery, respectively. Global (i.e., small-worldness) and local (i.e., clustering and strength) connectivity measures were calculated on a single-subject connectivity matrix of EEG data. Results: Both ELGA and late-preterm infants showed small-worldness organization at 35 weeks’ post-conception. The ELGA group had the strength parameter of the theta frequency band lower in the right than in the left hemisphere. This asymmetry did not emerge in the late-preterm group. Moreover, the mean strength parameter was significantly greater in the right hemisphere in the late preterms than in the ELGA group. Conclusion: EEG connectivity measures could represent an index of left-to-right maturation and developmental disadvantage in extremely preterm infants.