Brain Age Prediction in Neuroimaging

Brain age prediction employs computational models to estimate an individual’s biological brain age from neuroimaging data, most commonly structural magnetic resonance imaging. By comparing predicted and chronological age, the resulting brain age gap (BAG) serves as a surrogate marker of accelerated or decelerated brain ageing. This framework captures subtle, diffuse patterns of volume loss, microstructural change and functional reorganisation, offering insights into normal and pathological ageing. It has been applied across lifespan cohorts to chart healthy trajectories, in clinical populations to detect preclinical neurodegenerative changes, and in epidemiological studies to link lifestyle, metabolic and genetic factors to brain health. Integration of multimodal imaging, machine and deep learning techniques, and genetic analyses has enhanced predictive accuracy and biological interpretability, paving the way for personalised risk profiling and monitoring of intervention outcomes.

Research from Nature Portfolio

Recent studies have elucidated the genetic architecture underpinning distinct brain ageing phenotypes derived from grey matter, white matter and functional connectivity. Genome-wide analyses have uncovered multiple loci exerting differential effects on these modalities, with heritability enrichment in conserved genomic regions and cell-type specificity among oligodendrocytes and astrocytes. Employing Mendelian randomisation, investigators have proposed causal influences of chronic cardiometabolic and neurodegenerative conditions on accelerated brain ageing. In parallel, advanced deep learning frameworks trained on large structural MRI cohorts have refined age prediction accuracy and enabled discovery of sequence variants linked to cortical and white matter characteristics. These developments underscore the synergy of high-dimensional imaging, neural networks and genomics in revealing pathways of brain ageing with potential therapeutic implications.

Brain Age Prediction in Neuroimaging publication trend

The graph below shows the total number of articles in brain age prediction in neuroimaging across all publications each year (not limited to Nature Index journals).

Technical terms

Brain age gap (BAG): The numerical difference between an individual’s predicted brain age and their actual chronological age, indicative of accelerated or decelerated brain ageing.

Convolutional neural network: A deep learning model architecture that automatically learns spatial features from three-dimensional neuroimaging data to predict variables such as age.

Multimodal neuroimaging: The combined use of multiple imaging techniques—structural MRI, diffusion MRI and functional MRI—to capture complementary aspects of brain structure and function for age prediction.

References

1. Advanced structural brain aging in preclinical autosomal dominant Alzheimer disease. Molecular Neurodegeneration (2023).
2. The genetic architecture of multimodal human brain age. Nature Communications (2024).
3. Brain age prediction using deep learning uncovers associated sequence variants. Nature Communications (2019).
4. Ten Years of BrainAGE as a Neuroimaging Biomarker of Brain Aging: What Insights Have We Gained?. Frontiers in Neurology (2019).
5. Multimodality neuroimaging brain-age in UK biobank: relationship to biomedical, lifestyle, and cognitive factors. Neurobiology of Aging (2020).