ZusammenfassungThe deviation between chronological age and age as predicted from structural Magnetic Resonance Imaging (sMRI) data of the brain - the so-called Brain Age Gap – is increasingly used to quantify imaging deviations from normal aging and developmental trajectories. Here, we seek to develop, train, and validate a large-scale Brain Age model based on the data requested in this proposal using state-of-the-art machine learning algorithms. Most importantly, this would allow us to model the full range of brain structure variation in the population, providing the most comprehensive Brain Age model to date. Based on this model, we will then systematically investigate variables affecting the Brain Age Gap. Explicitly modeling confounding variables as well as interindividual differences will enable an estimation of Brain Age Gap variation in the population necessary for any potential clinical use of Brain Age Gap measures.

Schlüsselwörter-

EinrichtungenUniversität Münster, Institut für Epidemiologie der Universität Münster