ZusammenfassungCardiovascular diseases are highly prevalent in global civilization and account for the majority of deaths in Western societies. Recent demographic trends will make it difficult to hold the status quo with respect to cardiac morbidity and mortality. Established imaging techniques, particularly cardiac MRI, and its application in large-scale population-based cohort studies such as the NAKO have enormous potential for interdisciplinary research in this field – especially when leveraging fully automated analysis methods based on artificial deep neural networks. Data inferred from a corresponding imaging biobank will help to understand the transition from subclinical cardiac impairment to symptomatic cardiovascular disease and its association with both traditional and non-traditional risk factors. The identification of imaging biomarkers suitable for personalized risk stratification and early disease detection will possibly allow for novel intervention tactics prior to clinical disease manifestation. Our proposal accommodates these aims and structures them into a systematic workflow built around a proven, in-house developed deep learning algorithm for segmentation of cardiac MR image data. Utilizing the imaging parameters derived from these segmentations, we will focus on (sub)clinical disease progression and risk stratification as outlined above, as well as associations of ventricular remodeling with antineoplastic treatments such as thoracic radiotherapy, metabolic diseases such as type-2 diabetes, subclinical inflammation, and environmental or lifestyle influences such as air pollution, alcohol and coffee consumption or obstructive sleep apnea which will be par of follow-up proposals for data use. Upfront agreements have been made with other research groups to share data and collectively work on additional aims regarding the correlations of cardiac impairment with atherosclerosis, respiratory function, sarcopenia and aging-related brain alterations.

Schlüsselwörter-

EinrichtungenUniversitätsklinikum Freiburg, Helmholtz Zentrum München, Universitätsklinikum Heidelberg, Universität Tübingen, University of Freiburg, Deutsches Krebsforschungszentrum (DKFZ), Universitätsmedizin Rostock, University Hospital Heidelberg, Charité Universitätsmedizin Berlin, DKFZ, Universitätsklinikum Tübingen, Herr, Diagnostische und Interventionelle Radiologie, DKFZ Heidelberg, Radiologie/Uni Heidelberg, Klinikum der Universität München, Forschungszentrum Jülich