Automated cot-side tracking of functional brain age in preterm infants with routine EEG monitoring

A major challenge in the care of preterm infants is the early identification of compromised neurological development. While several measures are routinely used to track anatomical growth, there is a striking lack of reliable and objective tools for tracking maturation of early brain function; a cornerstone of lifelong neurological health. We present a cot-side method for measuring the functional maturity of the newborn brain based on routinely-available neurological monitoring with electroencephalography (EEG). We used a dataset EEG recordings from 65 infants to train a multivariable prediction of functional brain age (FBA) from EEG. Using machine learning on traditional and recently-developed computational EEG measures yielded an FBA that correlated strongly with the postmenstrual age of an infant. Moreover, individual babies follow well-defined individual trajectories. We validated the FBA predictor on independent data from a different site with different recording configuration. In a subgroup of infants with repeated EEG recordings, a persistently negative predicted age difference was associated with poor neurodevelopmental outcome. The FBA enables the tracking of functional neurodevelopment in preterm infants. Functional age assessment can be used to assist clinical management and identify infants who will benefit most from early intervention.