Multiscale entropy as a metric of brain maturation in a large cohort of typically developing children born preterm using longitudinal high-density EEG in the first two years of life

Abstract

Objective. We aimed to analyze whether complexity of brain electrical activity (EEG) measured by multiscale entropy (MSE) increases with brain maturation during the first two years of life. We also aimed to investigate whether this complexity shows regional differences across the brain, and whether changes in complexity are influenced by extrauterine life experience duration. Approach. We measured MSE of EEG signals recorded longitudinally using a high-density setup (64 or 128 electrodes) in 84 typically developing infants born preterm (<32 weeks’ gestation) from term age to two years. We analyzed the complexity index and maximum value of MSE over increasing age, across brain regions, and in function of extrauterine life duration, and used correlation matrices as a metric of functional connectivity of the cerebral cortex. Main results. We found an increase of strong inter-channel correlation of MSE (R>0.8) with increasing age. Regional analysis showed significantly increased MSE between 3 and 24 months of corrected age in the posterior and middle regions with respect to the anterior region. We found a weak relationship (adjusted R2=0.135) between MSE and extrauterine life duration. Significance. These findings suggest that brain functional connectivity increases with maturation during the first two years of life. EEG complexity shows regional differences with earlier maturation of the visual cortex and brain regions involved in joint attention than of regions involved in cognitive analysis, abstract thought, and social behavior regulation. Finally, our MSE analysis suggested only a weak influence of early extrauterine life experiences (prior to term age) on EEG complexity.

High‐density neonatal electroencephalography provides biomarkers for developmental trajectories

“…by using 64 or 128 channels). 1,2 If using appropriate mathematical tools, this has the potential to identify brain activities that were previously unknown or difficult to localize, and effectively characterize developmental neurophysiological trajectories in follow-up studies. Measures performed on high-density EEG signals might therefore be the next step when studying functional neuroplasticity induced by early intervention.…”

“…To enhance the spatial accuracy of EEG to physiologically relevant levels in particular in young infants, high‐density recordings are required (e.g. by using 64 or 128 channels) 1,2 . If using appropriate mathematical tools, this has the potential to identify brain activities that were previously unknown or difficult to localize, and effectively characterize developmental neurophysiological trajectories in follow‐up studies.…”

“…EEG can be recorded from the preterm period in newborn infants to document developmental patterns that reflect the progressive organization of brain networks. It can thus provide valuable biomarkers for early developmental trajectories 2 . For example, EEG can be used to identify biomarkers of attention, memory, language, and social cognition in infants and young children, which can be used to predict developmental outcomes and to identify children who are at risk for developmental delay or disorders.…”

“…This is also what children tell of their lived experience. 2 However, whether school completion is the most relevant parameter to evaluate educational achievement is disputable. Participation is increasingly becoming a goal in rehabilitation as well as in research.…”

Epilepsy is a continuing disease

“…For example, coherence measures reflect sharing of electrophysiological information across different brain regions. Entropy measures of the apparently random information contained in EEG signals also appear to be promising, including from a neurodevelopmental perspective 5 . The rapid development of machine learning will likely improve the speed and relevance of the large amounts of data to be processed.…”