AI-CVD-HFA

The AI-CVD-HF model introduces a novel imaging-based approach for heart failure (HF) risk prediction by combining AI-derived biomarkers extracted from Coronary Artery Calcium (CAC) scans with basic demographic factors, including age and sex. Compared with the baseline version of the recently developed PREVENT-HF risk score, the model demonstrated superior discrimination and calibration, while maintaining consistent predictive performance across different age groups, sexes, diverse racial and ethnic populations, and an external validation cohort. These findings highlight the model’s strong potential for generalizability and its clinical utility in opportunistic HF risk assessment using routine chest CT scans. Nevertheless, prospective randomized clinical trials and cost-effectiveness analyses are still required before widespread clinical implementation can be recommended.