Age is a significant risk factor for cardiovascular, neurodegenerative, oncologic and musculoskeletal disorders and an important patient-associated parameter affecting medical decisions in a clinical context. Due to the high inter-individual variation of age-related morphological and functional changes in individuals, the concept of biological age has been introduced aiming to describe the extent of these age-dependent changes in individuals. Definition of biological age is not clear and mostly relies on genetic, metabolic or functional parameters. Assessment of biological age using medical imaging – although providing potential advantages such as organ-specific BA estimation – however is still not fully investigated mainly due to lack of sufficient amount of standardized cross-sectional imaging data of the underlying population. The UK Biobank MR study provides a standardized data set of thousands of whole body MR data sets together with epidemiological, anthropomorphic, functional and laboratory parameters from the participating individuals.
The aim of this project is to derive organ-specific estimates for biological age based on whole body MRI data and to identify risk factors for accelerate biological aging.
To this end, target organs will be segmented on MR data sets using a deep neural approach. Image-based chronological age estimation will be implemented using a deep learning-based regression model. Subsequently, estimates for organ-specific biological age will be derived in a data-driven iterative approach.
Image-based estimates of biological age will be validated by comparison to biological age estimates from epidemiologic, anthropomorphic, laboratory and functional non-imaging data. Finally, risk-factors influencing biological age distribution will be identified using statistical methods.
