MaRS Cardiovascular Risk Score Non Clinical Validated

This preprint introduces the MaRS (Machine learning–augmented Risk Score), a non-laboratory cardiovascular risk model designed to be accessible to all healthcare professionals. The model builds upon traditional logistic regression while being extensible to machine learning, enabling the detection of new parameter interactions (e.g., BMI × diabetes, age × smoking) and the validation of additional predictors. MaRS fills the gap between existing non-lab scores (e.g., Framingham BMI-based, INTERHEART non-lab) and clinical practice, offering a hybrid, interpretable and evolving approach to cardiovascular risk prediction.