Graph-theoretical analysis of EEG functional brain network development in preterm infants

OBJECTIVE: To investigate preterm brain network maturation using electroencephalography connectivity and graph theory. METHODS: Sixty-one preterm infants (25-36 weeks gestational age) were studied, including 40 with longitudinal electroencephalography. Resting-state recordings across δ (0.5-4 Hz), θ (4-8 Hz), α (8-13 Hz), and β (13-30 Hz) bands were analyzed using weighted phase-lag index (wPLI), with false discovery rate correction. Graph metrics (small-worldness, efficiency, degree, betweenness, clustering) were computed across 27-50% sparsity. Developmental trajectories were modeled with generalized additive models and linear mixed-effects models. RESULTS: wPLI revealed 25 significant connections, dominated by interhemispheric frontopolar synchrony. θ-band clustering at F4 peaked at 36 weeks. Between 35-38 weeks, β- and θ-band networks showed significant declines, whereas δ-band connectivity demonstrated enhanced global integration and efficiency. Longitudinal models confirmed post-menstrual age (PMA)-dependent increases in parietal hub (P4) betweenness, degree, and nodal efficiency in the θ-band. CONCLUSION: Thirty-six weeks PMA marks a critical neurodevelopmental transition, characterized by δ-band hub emergence and parietal θ-band integration, offering potential biomarkers for preterm network maturation.