Principal Investigator: Dr Markus Ankenbrand
University Hospital Würzburg, Wurzburg, GermanyTags: 53911, Association, cardiovascular disease, genetics/genotyping, Imaging, Machine Learning, phenotype
Cardiovascular disease (CVD) is the leading cause of death in many industrial countries. As a complex disease beside well known risk factors many other factors including morphology, genetics, lifestyle and environment play an important role in disease genesis and progression. Yet, despite enormous research efforts not all factors and their intricate interactions are fully understood. In particular a large part of the heredibility of CVD remains unexplained. Even for many variants that are found to be associated with CVD risk their mechanism is unknown. A possible explanation for our inability to fully characterize the relationship between genetics and CVD is that CVD is a complex disease with many causations and sub-forms. This study aims to better characterize phenotypes related to CVD to uncover additional causal mutations and to link known genetic variants to specific phenotypic changes. In order to do that, state-of-the-art machine learning methods will be employed to automatically extract quantitative values from complex data like 4D cardiac magnetic resonance images. While this is a preliminary study aiming to explore the possibilities of refined phenotypes, a deeper understanding of the genetic background of CVD could help in better diagnosis, selection of the optimal treatment and the development of new therapeutic approaches.