EEG-based age-prediction models as stable and heritable indicators of brain maturational level in children and adolescents

Knowledge mapping of neonatal electroencephalogram: A bibliometric analysis (2004-2022)

BACKGROUND

Electroencephalography (EEG), a widely used noninvasive neurophysiological diagnostic tool, has experienced substantial advancements from 2004 to 2022, particularly in neonatal applications. Utilizing a bibliometric methodology, this study delineates the knowledge structure and identifies emergent trends within neonatal EEG research.

METHODS

An exhaustive literature search was conducted on the Web of Science Core Collection (WoSCC) database to identify publications related to neonatal EEG from 2004 to 2022. Analytical tools such as VOSviewer, CiteSpace, and the R package "bibliometrix" were employed to facilitate this investigation.

RESULTS

The search yielded 2501 articles originating from 79 countries, with the United States and England being the predominant contributors. A yearly upward trend in publications concerning neonatal EEG was observed. Notable research institutions leading this field include the University of Helsinki, University College London, and University College Cork. Clinical Neurophysiology is identified as the foremost journal in this realm, with Pediatrics as the most frequently co-cited journal. The collective body of work from 9977 authors highlights Sampsa Vanhatalo as the most prolific contributor, while Mark Steven Scher is recognized as the most frequently co-cited author. Key terms such as "seizures," "epilepsy," "hypoxic-ischemic encephalopathy," "amplitude-integrated EEG," and "brain injury" represent the focal research themes.

CONCLUSION

This bibliometric analysis offers the first comprehensive review, encapsulating research trends and progress in neonatal EEG. It reveals current research frontiers and crucial directions, providing an essential resource for researchers engaged in neonatal neuroscience.

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

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 analysed a 10- to 15-min segment of 18-channel EEG recorded during light sleep (N1 and N2 states).

FINDINGS

The FBA 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

A 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.

eXplainable Artificial Intelligence (XAI) in aging clock models

XAI is a rapidly progressing field of machine learning, aiming to unravel the predictions of complex models. XAI is especially required in sensitive applications, e.g. in health care, when diagnosis, recommendations and treatment choices might rely on the decisions made by artificial intelligence systems. AI approaches have become widely used in aging research as well, in particular, in developing biological clock models and identifying biomarkers of aging and age-related diseases. However, the potential of XAI here awaits to be fully appreciated. We discuss the application of XAI for developing the "aging clocks" and present a comprehensive analysis of the literature categorized by the focus on particular physiological systems.

Mapping neurodevelopment with sleep macro- and micro-architecture across multiple pediatric populations

Profiles of sleep duration and timing and corresponding electroencephalographic activity reflect brain changes that support cognitive and behavioral maturation and may provide practical markers for tracking typical and atypical neurodevelopment. To build and evaluate a sleep-based, quantitative metric of brain maturation, we used whole-night polysomnography data, initially from two large National Sleep Research Resource samples, spanning childhood and adolescence (total N = 4,013, aged 2.5 to 17.5 years): the Childhood Adenotonsillectomy Trial (CHAT), a research study of children with snoring without neurodevelopmental delay, and Nationwide Children's Hospital (NCH) Sleep Databank, a pediatric sleep clinic cohort. Among children without neurodevelopmental disorders (NDD), sleep metrics derived from the electroencephalogram (EEG) displayed robust age-related changes consistently across datasets. During non-rapid eye movement (NREM) sleep, spindles and slow oscillations further exhibited characteristic developmental patterns, with respect to their rate of occurrence, temporal coupling and morphology. Based on these metrics in NCH, we constructed a model to predict an individual's chronological age. The model performed with high accuracy (r = 0.93 in the held-out NCH sample and r = 0.85 in a second independent replication sample - the Pediatric Adenotonsillectomy Trial for Snoring (PATS)). EEG-based age predictions reflected clinically meaningful neurodevelopmental differences; for example, children with NDD showed greater variability in predicted age, and children with Down syndrome or intellectual disability had significantly younger brain age predictions (respectively, 2.1 and 0.8 years less than their chronological age) compared to age-matched non-NDD children. Overall, our results indicate that sleep architectureoffers a sensitive window for characterizing brain maturation, suggesting the potential for scalable, objective sleep-based biomarkers to measure neurodevelopment.

Application of bidirectional long short-term memory network for prediction of cognitive age

Electroencephalography (EEG) measures changes in neuronal activity and can reveal significant changes from infancy to adulthood concomitant with brain maturation, making it a potential physiological marker of brain maturation and cognition. To investigate a promising deep learning tool for EEG classification, we applied the bidirectional long short-term memory (BLSTM) algorithm to analyze EEG data from the pediatric EEG laboratory of Taipei Tzu Chi Hospital. The trained BLSTM model was 86% accurate when identifying EEGs from young children (8 months-6 years) and adolescents (12-20 years). However, there was only a modest classification accuracy (69.3%) when categorizing EEG samples into three age groups (8 months-6 years, 6-12 years, and 12-20 years). For EEG samples from patients with intellectual disability, the prediction accuracy of the trained BLSTM model was 46.4%, which was significantly lower than its accuracy for EEGs from neurotypical patients, indicating that the individual's intelligence plays a major role in the age prediction. This study confirmed that scalp EEG can reflect brain maturation and the BLSTM algorithm is a feasible deep learning tool for the identification of cognitive age. The trained model can potentially be applied to clinical services as a supportive measurement of neurodevelopmental status.

Optimization of time series features to estimate brain age in children from electroencephalography

Functional brain age measures in children, derived from the electroencephalogram (EEG), offer direct and objective measures in assessing neurodevelopmental status. Here we explored the effectiveness of 32 preselected 'handcrafted' EEG features in predicting brain age in children. These features were benchmarked against a large library of highly comparative multivariate time series features (>7000 features). Results showed that age predictors based on handcrafted EEG features consistently outperformed a generic set of time series features. These findings suggest that optimization of brain age estimation in children benefits from careful preselection of EEG features that are related to age and neurodevelopmental trajectory. This approach shows potential for clinical translation in the future.Clinical Relevance-Handcrafted EEG features provide an accurate functional neurodevelopmental biomarker that tracks brain function maturity in children.

High-Resolution and Multidimensional Phenotypes Can Complement Genomics Data to Diagnose Diseases in the Neonatal Population

Advances in genomic medicine have greatly improved our understanding of human diseases. However, phenome is not well understood. High-resolution and multidimensional phenotypes have shed light on the mechanisms underlying neonatal diseases in greater details and have the potential to optimize clinical strategies. In this review, we first highlight the value of analyzing traditional phenotypes using a data science approach in the neonatal population. We then discuss recent research on high-resolution, multidimensional, and structured phenotypes in neonatal critical diseases. Finally, we briefly introduce current technologies available for the analysis of multidimensional data and the value that can be provided by integrating these data into clinical practice. In summary, a time series of multidimensional phenome can improve our understanding of disease mechanisms and diagnostic decision-making, stratify patients, and provide clinicians with optimized strategies for therapeutic intervention; however, the available technologies for collecting multidimensional data and the best platform for connecting multiple modalities should be considered.

Social conformity is associated with inter-trial electroencephalogram variability

Human society encompasses diverse social influences, and people experience events differently and may behave differently under such influence, including in forming an impression of others. However, little is known about the underlying neural relevance of individual differences in following others' opinions or social norms. In the present study, we designed a series of tasks centered on social influence to investigate the underlying relevance between an individual's degree of social conformity and their neural variability. We found that individual differences under the social influence are associated with the amount of inter-trial electroencephalogram (EEG) variability over multiple stages in a conformity task (making face judgments and receiving social influence). This association was robust in the alpha band over the frontal and occipital electrodes for negative social influence. We also found that inter-trial EEG variability is a very stable, participant-driven internal state measurement and could be interpreted as mindset instability. Overall, these findings support the hypothesis that higher inter-trial EEG variability may be related to higher mindset instability, which makes participants more vulnerable to exposed external social influence. The present study provides a novel approach that considers the stability of one's endogenous neural signal during tasks and links it to human social behaviors.

Heritability of REM sleep neurophysiology in adolescence

Alterations of rapid eye movement (REM) sleep have long been observed in patients with psychiatric disorders and proposed as an endophenotype-a link between behavior and genes. Recent experimental work has shown that REM sleep plays an important role in the emotional processing of memories, emotion regulation, and is altered in the presence of stress, suggesting a mechanism by which REM sleep may impact psychiatric illness. REM sleep shows a developmental progression and increases during adolescence-a period of rapid maturation of the emotional centers of the brain. This study uses a behavioral genetics approach to understand the relative contribution of genes, shared environmental and unique environmental factors to REM sleep neurophysiology in adolescents. Eighteen monozygotic (MZ; n = 36; 18 females) and 12 dizygotic (DZ; n = 24; 12 females) same-sex twin pairs (mean age = 12.46; SD = 1.36) underwent whole-night high-density sleep EEG recordings. We find a significant genetic contribution to REM sleep EEG power across frequency bands, explaining, on average, between 75 to 88% of the variance in power, dependent on the frequency band. In the lower frequency bands between delta and sigma, however, we find an additional impact of shared environmental factors over prescribed regions. We hypothesize that these regions may reflect the contribution of familial and environmental stress shared amongst the twins. The observed strong genetic contribution to REM sleep EEG power in early adolescence establish REM sleep neurophysiology as a potentially strong endophenotype, even in adolescence-a period marked by significant brain maturation.

Electroencephalographic studies in growth-restricted and small-for-gestational-age neonates

Foetal growth restriction (FGR) and being born small for gestational age (SGA) are associated with neurodevelopmental delay. Early diagnosis of neurological damage is difficult in FGR and SGA neonates. Electroencephalography (EEG) has the potential as a tool for the assessment of brain development in FGR/SGA neonates. In this review, we analyse the evidence base on the use of EEG for the assessment of neonates with FGR or SGA. We found consistent findings that FGR/SGA is associated with measurable changes in the EEG that present immediately after birth and persist into childhood. Early manifestations of FGR/SGA in the EEG include changes in spectral power, symmetry/synchrony, sleep-wake cycling, and the continuity of EEG amplitude. Later manifestations of FGR/SGA into infancy and early childhood include changes in spectral power, sleep architecture, and EEG amplitude. FGR/SGA infants had poorer neurodevelopmental outcomes than appropriate for gestational age controls. The EEG has the potential to identify FGR/SGA infants and assess the functional correlates of neurological damage. IMPACT: FGR/SGA neonates have significantly different EEG activity compared to AGA neonates. EEG differences persist into childhood and are associated with adverse neurodevelopmental outcomes. EEG has the potential for early identification of brain impairment in FGR/SGA neonates.

Large-scale collaboration in ENIGMA-EEG: A perspective on the meta-analytic approach to link neurological and psychiatric liability genes to electrophysiological brain activity

BACKGROUND AND PURPOSE

The ENIGMA-EEG working group was established to enable large-scale international collaborations among cohorts that investigate the genetics of brain function measured with electroencephalography (EEG). In this perspective, we will discuss why analyzing the genetics of functional brain activity may be crucial for understanding how neurological and psychiatric liability genes affect the brain.

METHODS

We summarize how we have performed our currently largest genome-wide association study of oscillatory brain activity in EEG recordings by meta-analyzing the results across five participating cohorts, resulting in the first genome-wide significant hits for oscillatory brain function located in/near genes that were previously associated with psychiatric disorders. We describe how we have tackled methodological issues surrounding genetic meta-analysis of EEG features. We discuss the importance of harmonizing EEG signal processing, cleaning, and feature extraction. Finally, we explain our selection of EEG features currently being investigated, including the temporal dynamics of oscillations and the connectivity network based on synchronization of oscillations.

RESULTS

We present data that show how to perform systematic quality control and evaluate how choices in reference electrode and montage affect individual differences in EEG parameters.

CONCLUSION

The long list of potential challenges to our large-scale meta-analytic approach requires extensive effort and organization between participating cohorts; however, our perspective shows that these challenges are surmountable. Our perspective argues that elucidating the genetic of EEG oscillatory activity is a worthwhile effort in order to elucidate the pathway from gene to disease liability.

The association of polygenic risk for schizophrenia, bipolar disorder, and depression with neural connectivity in adolescents and young adults: examining developmental and sex differences

Neurodevelopmental abnormalities in neural connectivity have been long implicated in the etiology of schizophrenia (SCZ); however, it remains unclear whether these neural connectivity patterns are associated with genetic risk for SCZ in unaffected individuals (i.e., an absence of clinical features of SCZ or a family history of SCZ). We examine whether polygenic risk scores (PRS) for SCZ are associated with functional neural connectivity in adolescents and young adults without SCZ, whether this association is moderated by sex and age, and if similar associations are observed for genetically related neuropsychiatric PRS. One-thousand four-hundred twenty-six offspring from 913 families, unaffected with SCZ, were drawn from the Collaborative Study of the Genetics of Alcoholism (COGA) prospective cohort (median age at first interview = 15.6 (12-26), 51.6% female, 98.1% European American, 41% with a family history of alcohol dependence). Participants were followed longitudinally with resting-state EEG connectivity (i.e., coherence) assessed every two years. Higher SCZ PRS were associated with elevated theta (3-7 Hz) and alpha (7-12 Hz) EEG coherence. Associations differed by sex and age; the most robust associations were observed between PRS and parietal-occipital, central-parietal, and frontal-parietal alpha coherence among males between ages 15-19 (B: 0.15-0.21, p < 10^-4). Significant associations among EEG coherence and Bipolar and Depression PRS were observed, but differed from SCZ PRS in terms of sex, age, and topography. Findings reveal that polygenic risk for SCZ is robustly associated with increased functional neural connectivity among young adults without a SCZ diagnosis. Striking differences were observed between men and women throughout development, mapping onto key periods of risk for the onset of psychotic illness and underlining the critical importance of examining sex differences in associations with neuropsychiatric PRS across development.

The Speed of Development of Adolescent Brain Age Depends on Sex and Is Genetically Determined

Children and adolescents show high variability in brain development. Brain age-the estimated biological age of an individual brain-can be used to index developmental stage. In a longitudinal sample of adolescents (age 9-23 years), including monozygotic and dizygotic twins and their siblings, structural magnetic resonance imaging scans (N = 673) at 3 time points were acquired. Using brain morphology data of different types and at different spatial scales, brain age predictors were trained and validated. Differences in brain age between males and females were assessed and the heritability of individual variation in brain age gaps was calculated. On average, females were ahead of males by at most 1 year, but similar aging patterns were found for both sexes. The difference between brain age and chronological age was heritable, as was the change in brain age gap over time. In conclusion, females and males show similar developmental ("aging") patterns but, on average, females pass through this development earlier. Reliable brain age predictors may be used to detect (extreme) deviations in developmental state of the brain early, possibly indicating aberrant development as a sign of risk of neurodevelopmental disorders.

Environmental Factors Shape Sleep EEG Connectivity During Early Adolescence

Quantifying the degree to which genetic and environmental factors shape brain network connectivity is critical to furthering our understanding of the developing human brain. Sleep, a state of sensory disengagement, provides a unique opportunity to study brain network activity noninvasively by means of sleep electroencephalography (EEG) coherence. We conducted a high-density sleep EEG study in monozygotic (MZ; n = 38; mean age = 12.46; 20 females) and dizygotic (DZ; n = 24; mean age = 12.50; 12 females) twins to assess the heritability of sleep EEG coherence in early adolescence—a period of significant brain rewiring. Structural equation modeling was used to estimate three latent factors: genes, environmental factors shared between twins and environmental factors unique to each twin. We found a strong contribution of unique environmental factors (66% of the variance) and moderate genetic influence (19% of the variance) on sleep EEG coherence across frequencies and sleep states. An exception to this was sleep spindle activity, an index of the thalamocortical network, which showed on average a genetic contribution of 48% across connections. Furthermore, we observed high intraindividual stability of coherence across two consecutive nights suggesting that despite only a modest genetic contribution, sleep EEG coherence is like a trait. Our findings in adolescent humans are in line with earlier findings in animals that show the primordial cerebral map and its connections are plastic and it is through interaction with the environment that the pattern of brain network connectivity is shaped. Therefore, even in twins living together, small differences in the environment may cascade into meaningful differences in brain connectivity.

EEG-based age-prediction models as stable and heritable indicators of brain maturational level in children and adolescents

The human brain shows remarkable development of functional brain activity from childhood to adolescence. Here, we investigated whether electroencephalogram (EEG) recordings are suitable for predicting the age of children and adolescents. Moreover, we investigated whether overestimation or underestimation of age was stable over longer time periods, as stable prediction error can be interpreted as reflecting individual brain maturational level. Finally, we established whether the age-prediction error was genetically determined. Then, 3 min eyes-closed resting-state EEG data from the longitudinal EEG studies of Netherlands Twin Register (NTR; n = 836) and Washington University in St. Louis (n = 702) were used at ages 5, 7, 12, 14, 16, and 18. Longitudinal data were available within childhood (5-7 years) and adolescence (16-18 years). We calculated power in 1 Hz wide bins (1-24 Hz). Random forest (RF) regression and relevance vector machine with sixfold cross-validation were applied. The best mean absolute prediction error was obtained with RF (1.22 years). Classification of childhood versus puberty/adolescence reached over 94% accuracy. Prediction errors were moderately to highly stable over periods of 1.5-2.1 years (0.53 < r < 0.74) and signifcantly affected by genetic factors (heritability between 42 and 79%). Our results show that age prediction from low-cost EEG recordings is comparable in accuracy to those obtained with magnetic resonance imaging. Children and adolescents showed stable overestimation or underestimation of their age, which means that some participants have stable brain activity patterns that reflect those of an older or younger age, and could therefore reflect individual brain maturational level. This prediction error is heritable, suggesting that genes underlie maturational level of functional brain activity. We propose that age prediction based on EEG recordings can be used for tracking neurodevelopment in typically developing children, in preterm children, and in children with neurodevelopmental disorders.