7 October 2022 in bioRxiv
Normative modelling is an emerging technique for parsing heterogeneity in clinical cohorts. This can be implemented in practice using hierarchical Bayesian regression, which provides an elegant probabilistic solution to handle site variation in a federated learning framework. However, applications of this method to date have employed a Gaussian assumption, which may be restrictive in some applications. We have extended the hierarchical Bayesian regression framework to flexibly model non-Gaussian data with heteroskdastic skewness and kurtosis. To this end, we employ a flexible distribution from the sinh-arcsinh (SHASH) family, and introduce a novel reparameterisation and a Markov chain Monte Carlo sampling approach to perform inference in this model. Using a large neuroimaging dataset collected at 82 different sites, we show that the results achieved with this extension are equivalent or better than a warped Bayesian linear regression baseline model on most datasets, whilst providing better control over the parameters governing the shape of distributions that approach is able to model. We also demonstrate that the attained flexibility is useful for accurately modelling highly nonlinear relationships between aging and imaging derived phenotypes, which shows that the extension is important for pushing the field of normative modelling forward. All methods described here are available in the open-source pcntoolkit . Highlights: We extended the Hierarchical Bayesian Regression framework for normative modelling Our extension allows modelling data with heteroskedastic skewness and kurtosis We developed a reparameterization of the SHASH distribution, suitable for sampling We provide the first implementation of the SHASH distribution in a fully Bayesian framework Results show that the extension outperforms current methods on various measures
17 March 2026 in Communications Biology
Normative modeling provides a principled framework for quantifying individual deviations from typical brain development and is increasingly used to study heterogeneity in neuropsychiatric conditions. While widely applied to structural phenotypes, functional normative models remain underdeveloped. Here, we introduce MEGaNorm, a normative modeling framework for charting lifespan trajectories of resting-state magnetoencephalography (MEG) brain oscillations. Using a large, multi-site dataset comprising 1846 individuals aged 6-88 and spanning three MEG systems, we model relative oscillatory power in canonical frequency bands using hierarchical Bayesian regression, accounting for age, sex, and site effects. To support interpretation at multiple scales, we introduce Neuro-Oscillo Charts, visual tools that summarize normative trajectories at the population level and quantify individual-level deviations, enabling personalized assessment of functional brain dynamics. Applying this framework to a Parkinson’s disease cohort (n = 160), we demonstrate that normative deviation scores reveal disease-related abnormalities and identify a continuum of patients in the theta-beta deviation space. This work establishes a multi-site normative reference for resting-state MEG oscillations (3-40 Hz) across a broad age range, enabling population-level characterization and individualized benchmarking. All models and tools are openly available and designed for federated, continual adaptation as new data become available, providing a methodological foundation toward precision neuropsychiatry.