Harmonized-Multinational qEEG norms (HarMNqEEG).

Area of Science:

• Neuroscience
• Biomedical Engineering
• Medical Imaging

Background:

• Quantitative electroencephalography (qEEG) methods have limitations in detecting brain developmental disorders due to a lack of cross-spectral information and geographical diversity.
• Previous qEEG studies often omitted crucial functional connectivity descriptors and failed to account for site-specific variations, leading to increased nuisance variance.

Purpose of the Study:

• To develop higher-sensitivity qEEG methods for detecting brain developmental disorders by integrating cross-spectral information and addressing geographical variance.
• To create lifespan Riemannian multinational qEEG norms for cross-spectral tensors to harmonize data and reduce bias.

Main Methods:

• Utilized data from 1564 subjects across 9 countries and 12 devices through the HarMNqEEG project.
• Calculated lifespan Riemannian norms for cross-spectral tensors, accounting for developmental trajectories using additive mixed-effects models.
• Developed methods to demonstrate and correct for qEEG "batch effects" to compute harmonized z-scores.

Main Results:

• Established harmonized Riemannian norms from multinational qEEG data, enabling the calculation of standardized z-scores.
• Demonstrated that harmonized Riemannian norms increase diagnostic accuracy in predicting brain dysfunction from malnutrition and COVID-19.
• Showcased the utility of these norms in identifying early-life malnutrition and COVID-19-induced brain dysfunction.

Conclusions:

• The developed harmonized Riemannian qEEG norms offer a significant advancement in detecting brain developmental disorders and dysfunction.
• This work contributes to creating bias-free, low-cost neuroimaging technologies applicable across diverse healthcare settings.
• Open-sourced code and data facilitate the calculation of individual z-scores, promoting wider research and clinical application.