Linked projectsNAKO-727

SummaryMetabolically vulnerable individuals, e.g. with prevalent diabetes or hypertension, are at higher risk for cardiovascular disease (CVD), and have higher susceptibility for the unfavorable impact of environmental risk factors. Personalized markers of early CVD tailored to the metabolic phenotype thus have the potential to improve risk stratification and enable early individualized intervention. The current proposal aims to use MRI data of the GNC to characterize left ventricular (LV) wall thickness and right ventricular (RV) pathologies in metabolically vulnerable individuals and assess the impact of envi-ronmental factors on this cardiac remodeling. We aim to use AI-supported algorithms to identify RV pathologies and derive a segmentation of the LV wall according to the 16-segment model of the American Heart Association. Metabolically vulnerable pheno-types are defined by combinations of hypertension, obesity, hyperlipidemia and diabetes. Cardiac remodel-ing will be characterized for each phenotype, and compared between groups. Using representation learn-ing, radiomics features indicative of each phenotype will be assessed. We then evaluate the association of environmental risk factors, such as air temperature and air pollution, with cardiac remodeling and evaluate the role of metabolic phenotype by causal mediation analysis.

Keywords-

InstitutionsUniversitätsklinikum Freiburg, Helmholtz Zentrum München