Predicting coronary artery calcification from demographic, clinical and lifestyle data: a machine learning-based approach

Cardiovascular disease is the most common cause of death worldwide and causes an immense burden on patient morbidity, healthcare resources and budgets. Currently established modifiable risk factors explain approximately 50% of the risk of incident cardiovascular disease (Magnussen C et al. N Engl J Med 2023;389:1273-1285). This project aims to elucidate at least in part factors included in the remaining half. While the contribution of CAC to the risk of cardiovascular disease has been established, CT scans required to measure CAC scores are expensive and burdensome. 
The proposed project aims to lower the threshold of estimating CAC by developing an open-source machine learning algorithm based on demographic, lifestyle and clinical chemistry data. Successfully completing the proposed project contributes to healthy ageing by establishing an inexpensive and non-invasive method of measuring CAC. In clinical care, information obtained by the model could be used to initiate risk stratification and personalized prevention and treatment strategies for patients at risk of vascular calcification and cardiovascular disease. In addition, modifiable risk factors such as lifestyle factors could be further addressed in follow-up intervention studies to improve clinical outcomes. The ultimate goal is not only to lower the risk of premature mortality, but also to reduce disease burden due to cardiovascular disease, in order to add more healthy years to the population