Cardiovascular calcification: artificial intelligence and big data accelerate mechanistic discovery

Abstract

Cardiovascular calcification is a health disorder with increasing prevalence and high morbidity and mortality. The only available therapeutic options for calcific vascular and valvular heart disease are invasive transcatheter procedures or surgeries that do not fully address the wide spectrum of these conditions; therefore, an urgent need exists for medical options. Cardiovascular calcification is an active process, which provides a potential opportunity for effective therapeutic targeting. Numerous biological processes are involved in calcific disease, including matrix remodelling, transcriptional regulation, mitochondrial dysfunction, oxidative stress, calcium and phosphate signalling, endoplasmic reticulum stress, lipid and mineral metabolism, autophagy, inflammation, apoptosis, loss of mineralization inhibition, impaired mineral resorption, cellular senescence and extracellular vesicles that act as precursors of microcalcification. Advances in molecular imaging and big data technology, including in multiomics and network medicine, and the integration of these approaches are helping to provide a more comprehensive map of human disease. In this Review, we discuss ectopic calcification processes in the cardiovascular system, with an emphasis on emerging mechanistic knowledge obtained through patient data and advances in imaging methods, experimental models and multiomics-generated big data. We also highlight the potential and challenges of artificial intelligence, machine learning and deep learning to integrate imaging and mechanistic data for drug discovery.