Agatston-2.0: AI-Derived Calcium Burden and Plaque Density Profiling Improve CHD Risk Stratification Within CAC Scores 1–99

Can AI turn the CAC 1–99 “gray zone” into actionable cardiovascular risk?

A new HeartLung.AI study, to be presented at ESC Congress 2026 in Munich, explores how Agatston-2.0 can improve coronary heart disease (CHD) risk stratification among individuals with conventional CAC scores of 1–99.

The study analyzed 1,542 participants from MESA and the Framingham Heart Study, with a median follow-up of 12.5 years and 151 CHD events.

Using measurements derived from a single CAC scan, Agatston-2.0 goes beyond the conventional calcium score by profiling three key characteristics of coronary calcification: relative plaque density, plaque distribution, and hyperdensity.

Key Findings

The results demonstrate substantial differences in risk within a group that would traditionally be classified together as CAC 1–99:

• 33% of participants were up-classified, accounting for 52% of 10-year CHD events.

• 41% were down-classified to a <5% 10-year CHD risk, accounting for only 27% of CHD events.

• The approach achieved a +34.8% net reclassification using information from the same CAC scan.

  
  

These findings suggest that the amount of calcium alone may not tell the whole story. How coronary calcium is distributed and how dense it is may provide additional information for distinguishing higher- and lower-risk individuals within the CAC 1–99 range.

One CAC range. Very different risk profiles. One scan with more actionable information.

📍 To be presented at ESC Congress 2026 · Munich