Agaston 2.0, A next-generation AI-based coronary calcium quantification approach to improve risk stratification among individuals with zero Agatston scores – Part I
- 3 days ago
- 2 min read
ESC Congress 2026 - 28 - 31 Aug 2026
Munich, Germany
Organised by: European Society of Cardiology (ESC) visual
Is a coronary calcium score of zero truly zero?
Presented at ESC Congress 2026, this HeartLung.AI study introduces Agatston-2.0, a next-generation AI-based approach designed to detect subtle coronary calcification that may be missed by conventional Agatston scoring.
Traditional Agatston scoring has transformed cardiovascular risk assessment over the past three decades. Yet its technical constraints—including the 130 HU threshold, minimum lesion size, 2.5–3 mm slice thickness, and peak-voxel intensity scoring—can leave very small or lower-density calcifications undetected.
Agatston-2.0 addresses these limitations using the same non-contrast CT scan, with no additional radiation. The framework combines deep-learning coronary segmentation, intensity calibration and noise normalization, continuous voxel-level density weighting, and spatial filtering to quantify subtle calcification within the coronary arteries.
Key Findings
The study analyzed 3,965 participants with conventional CAC = 0 from MESA and the Framingham Heart Study.
21.7% — 862 participants — had detectable AI-CAC > 0, meaning roughly one in five conventional “zeros” was not actually zero with Agatston-2.0.
After adjustment for traditional cardiovascular risk factors, AI-CAC > 0 was associated with:
73% higher risk of myocardial infarction
85% higher risk of hard coronary heart disease
71% higher risk of all coronary heart disease
Higher PREVENT ASCVD and overall ASCVD risk
Greater 10-year progression to conventional CAC > 0: 66.3% vs. 43.4%
The findings were also consistent across most clinically relevant subgroups, including age, sex, ethnicity, scanner type, hypertension, and dyslipidemia.
When AI-CAC was combined with additional AI-CVD imaging biomarkers extracted from the same CT, predicted 10-year cardiovascular risk within the conventional CAC-zero population ranged from just 0.3% to 15.7%—demonstrating substantial heterogeneity hidden behind a traditional score of zero.
The Takeaway
Agatston-2.0 refines the “power of zero”—it does not refute it.
Patients with no detectable calcification even by Agatston-2.0 represented a particularly low-risk subset, while AI-detected sub-threshold calcification identified individuals with meaningfully higher long-term cardiovascular risk.
Same scan. No extra radiation. A more precise understanding of cardiovascular risk.
Download the full ESC Congress 2026 presentation to explore the Agatston-2.0 methodology, clinical outcomes, subgroup analyses, and future directions.
Comments