Next-generation EEG methods for improving diagnostic and prognostic monitoring of child brain health

Developing reliable biomarkers to effectively measure child brain health, particularly in neurodevelopmental disorders, is a key challenge highlighted by the World Health Organization (WHO). Recently, the WHO has recognized electroencephalography (EEG) as a promising solution due to its affordability, non-invasive nature, and capacity for objective monitoring across all childhood ages. This talk presents our research on advancing EEG-based diagnostic and prognostic methods, from infancy to adolescence, utilizing machine learning (ML), artificial intelligence (AI) algorithms and computational modelling. Our approach leverages the flexibility of EEG to derive objective markers, ranging from low-cost, limited (2-channel) EEG for predicting brain age to the combined use of quantitative EEG (qEEG) and connectomics with age-binned head models applied to clinical (19-channel) and higher-density (128-channel) EEG configurations. We will also discuss the challenges involved in developing these diagnostic and prognostic algorithms in this population, including training, validation, and site harmonization confounders that are crucial for ensuring accuracy and generalizability. By employing advanced ML/AI tools, our goal is to create scalable solutions that enhance current diagnostic capabilities and enable proactive, personalized healthcare for children. Our efforts aim to pave the way for future innovations in child brain health monitoring, enabling objective, cost-effective EEG assessments on a larger scale, ultimately improving outcomes for pediatric populations, including those with neurodevelopmental disorders.

Reference: [1] Iyer et al. 2024. EBioMedicine; [2] Slater et al. 2023. The Lancet Digital Health; [3] Iyer et al. 2023. 45th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC); [4] Stevenson et al. 2017. Scientific reports; [5] Stevenson et al. 2020. Annals of clinical and translational neurology