Latent representations of early brain development: A multivariate normative model of brain structure and behaviour.

Individual variation in neurodevelopment plays a central role in shaping cognitive abilities and behavioural profiles, influencing both typical functioning and risk for neurodevelopmental conditions. While much research has focused on characterising trajectories of brain structure during development, this typically entails assessing brain regions individually, overlooking the multivariate nature of neuroimaging data. In this study, we trained an autoencoder to map latent representations of children's brain development using T1-weighted MRI scans from a paediatric cohort (n = 564, 55% male, mean age 4.90 years, range [0.11,15.38]). The latent representation from this model effectively captured demographic variables (age and sex), while preserving both global and local structural features. The model accurately reconstructed the data (mean reconstruction error = 0.04 ± 0.01) and captured demographic features with sex classification accuracy of 84% ± 4% and mean absolute error of 0.79 ± 0.06 years for age prediction, highlighting its sensitivity to developmental changes. We further validated the approach using correlation analysis showing that deviations from the latent norms were significantly associated with multiple cognitive and behavioural measures, suggesting that structural variations may reflect individual differences in neurodevelopment. Finally, we generated reference brain images that represent typical development and used them to visualise structural differences in individuals who deviate from this normative pattern. Our findings demonstrate that conditioned autoencoders, combined with multivariate normative modelling, offer a framework for characterising neurodevelopmental trajectories. This approach can identify meaningful deviations and has potential future applications across the lifespan.