Effective connectivity in long-term EEG monitoring in preterm infants

Abstract

Objective: Electroencephalogram (EEG) is commonly used to monitor brain activity in preterm neonates. Long-term EEG is used to predict outcomes by total band power. This study reports effective connectivity analyses in the fronto-parietal network in preterm infants.

Method: Directed transfer functions (DTFs) were used in fully automated calculations of effective connectivity. Continuous eight-channel EEG recordings were performed for three full days starting at a median of 5.3h (2.5-11.2h) after birth. Analyses were performed using the channels Fp1, Fp2, P3 and P4. DTF were calculated every second in every connection and channel for all frequencies from 1 Hz to 30 Hz. The median connectivity was used in the connectivity assessment.

Results: Parieto-frontal connectivity was significantly higher (0.22) than fronto-parietal connectivity (0.05). There was a significant decrease in parieto-frontal connectivity between day one (0.18) and three. Conversely, there was a significant increase in fronto-parietal connectivity from day one to day three (0.07).

Conclusion: Significant changes are seen in connectivity from day one to three, but there were no certain connectivity differences between gestational ages.

Significance: A rapid change in fronto-parietal connectivity was seen from day one to three which may indicate fast postpartum maturation.

Keywords: Connectivity; Default mode network; Developmental outcome; EEG; Maturation; Preterm.