A growth chart of brain function from infancy to adolescence based on electroencephalography

Background In children, objective, quantitative tools that determine functional neurodevelopment are scarce and rarely scalable for clinical use. Direct recordings of cortical activity using routinely acquired electroencephalography (EEG) offer reliable measures of brain function. Methods We developed and validated a measure of functional brain age (FBA) using a residual neural network-based interpretation of the paediatric EEG. In this cross-sectional study, we included 1056 children with typical development ranging in age from 1 month to 18 years. We analyzed a 10 to 15 minute segment of 18-channel EEG recorded during light sleep (N1 and N2 states). Findings The FBA obtained from EEG had a weighted mean absolute error (wMAE) of 0.85 years (95%CI: 0.69-1.02; n = 1056). A two-channel version of the FBA had a wMAE of 1.51 years (95%CI: 1.30-1.73; n = 1056) and was validated on an independent set of EEG recordings (wMAE = 2.27 years, 95%CI: 1.90-2.65; n = 723). Group-level maturational delays were also detected in a small cohort of children with Trisomy 21 (Cohen’s d = 0.36, p = 0.028). Interpretation An FBA, based on EEG, is an accurate, practical and scalable automated tool to track brain function maturation throughout childhood with accuracy comparable to widely used physical growth charts. Funding This research was supported by the National Health and Medical Research Council, Australia, Helsinki University Diagnostic Center Research Funds, Finnish Academy, Finnish Paediatric Foundation, and Sigrid Juselius Foundation. Evidence before this study Tools for objectively tracking neurodevelopment in paediatric populations using direct measurement of the brain are rare. Prior to conducting this study, we explored multiple databases (Google Scholar, PubMed, Web of Science) with search strategies that combined one or more of the terms “paediatric brain development”, “brain age”, “age estimation”, “MRI measurements”, “EEG measurements”, “machine learning”, “artificial intelligence”, “advanced ageing”, “neurodevelopmental delays” and “growth charts” with no restrictions on language and dates. In screening over 500 publications, 7 studies evaluated brain age in children using MRI and only a single study investigated maturation in EEG activity across discrete age bins. Added value of this study We formulated a measure of functional brain age (FBA) using state-of-the-art machine learning (ML) algorithms trained on a large, unique database consisting of multichannel clinical EEG recorded from N1/N2 sleep (n = 1056 children; 1 month to 17 years), with typical neurodevelopment confirmed at a 4-year follow-up. The FBA showed a high correlation with age and detected group-level differences associated with conditions of neurodevelopmental delay. Implications of all the available evidence Age is prominent within EEG recordings of N1/N2 sleep and is readily extracted using ML. Public release of the FBA estimator and the use of EEG, commonly delivered in outpatient settings, as the basis of age prediction enables clear translation of measures of ‘brain age’ to the clinic. Future work on EEG datasets across various neurodevelopmental profiles will enhance generalisability and user confidence in the clinical application of brain age. ### Competing Interest Statement J.A.R and S.V. hold a licensed patent on the burst metrics used in this paper. J.A.R. declares grants received from The Margaret Pemberton Foundation and the receipt of EEG equipment from Cadwell Industries. J.C. declares grants received from the following funding bodies: Medical Research Fund, National Health and Medical Research Council, Children's Hospital Foundation Fellowship. J.C. acknowledges paid lectures for the Sleep Health Foundation. K.K.I., A.K., M.W., S.J.V., L.L, L.M.H. and N.J.S declare no competing interests.