Artificial Intelligence and Cardiovascular Risk Assessment

RATIONALE

Atherosclerotic cardiovascular disease (ASCVD) remains the leading global cause of mortality, despite substantial advances in prevention, risk factors control, and pharmacological

therapy. Traditional risk prediction models, largely derived from population-based cohorts

and based on a limited set of classical well known risk factors, have provided an important

framework for estimating cardiovascular risk. However, while useful at the population level,

these models fall short when applied to individual patients. They may inadequately capture

heterogeneity across age groups, sex, ethnicity, and genetic background, and frequently

fail to account for lifetime exposure and dynamic changes in risk, thereby contributing to

persistent residual cardiovascular burden.

In recent years, the rapid expansion of large-scale biobanks, electronic health records, advanced imaging modalities, and multi-omics platforms has profoundly transformed the

landscape of cardiovascular risk assessment. These resources provide an unprecedented

opportunity to move beyond static and simplified risk estimation toward a more comprehensive and dynamic understanding of disease processes. When combined with artificial intelligence (AI) and machine learning approaches, these datasets enable the integration of complex, multidimensional information, capturing interactions between biological,

environmental, and behavioural determinants that often are not accessible through conventional models.

Such approaches hold the potential to redefine cardiovascular prevention by enabling more

precise, individualized risk stratification, earlier identification of high-risk phenotypes, and

more targeted therapeutic interventions. At the same time, they may help to better characterize subclinical disease, refine treatment thresholds, and optimize allocation of healthcare resources.

However, important challenges remain. These include the interpretability and transparency of AI-driven models, the need for rigorous validation across diverse populations and healthcare systems, issues related to data quality and bias, and the practical integration of

these tools into clinical workflows. Ethical considerations, including data governance and

patient privacy, also represent critical aspects that must be addressed to ensure responsible implementation.

This workshop in Fiesole is designed to address these opportunities and challenges by

combining high-level plenary lectures with interactive, hands-on sessions. The program

will provide participants with both a strong conceptual framework and practical skills, fostering a multidisciplinary dialogue between cardiology, data science, and public health.

The ultimate aim is to empower clinicians and researchers to critically evaluate, interpret,

and apply big data and AI tools in real-world settings, thereby contributing to more effective

and sustainable cardiovascular prevention strategies.