Specific changes in organ phenotypes such as degenerative changes of the spine, increased volumes of adipose tissue or diameter changes of large blood vessels may be indicative for chronic diseases and are associated with increasing age. However, as we all age at different rates our chronological age does not necessarily reflect our true biological age. Therefore, the proposed study will develop and test artificial intelligence-based models to quantify organ/tissue phenotypes and measures of biological age from MRI imaging data and will explore whether these measures can improve clinical guidelines for prevention of cardiovascular disease and cancer. Medical imaging may be an effective way to quantify organ and tissue phenotypes associated with accelerated aging by measuring anatomical changes visible in the image, which have not yet become clinically symptomatic. Unlike molecular measures (blood tests, genetic risk scores), image-based reflections of certain high risk phenotypes and aging can be calculated opportunistically using data acquired during daily clinical care, which are not routinely quantified due to limited resources and the lack of reliable methods. The deep learning-based methods developed in this study to quantify organ phenotypes and estimates of aging will address this unmet need and may help to personalize clinical decision-making for chronic disease screening and prevention.
