Coronary Artery Calcium Scoring: Agatston & Dose

Q: How much radiation does a calcium scoring CT deliver?

A dedicated calcium scoring CT is a low-dose exam. Typical effective doses are on the order of about 1 mSv with modern prospective ECG-triggered protocols, though the exact value depends on the scanner, patient size, and technique. It is far lower than a contrast-enhanced coronary CT angiogram.

Q: Why does the CT protocol matter so much for calcium scoring?

The Agatston score depends on a fixed 130 HU threshold, so any drift in CT number calibration, changes in slice thickness, tube voltage, or reconstruction can shift the score across a risk category. Standardized acquisition — 120 kVp, 2.5–3 mm slices, ECG gating, and verified HU calibration — is what makes scores comparable over time and between scanners.

Q: How reproducible is the calcium score?

Repeat calcium scores on the same patient typically vary by roughly 10–15% for scores above about 100, but variability is much larger for very low scores. This interscan variability limits how confidently small changes can be called real progression, and it is one reason serial scoring for individual patients must be interpreted cautiously.

Q: What is CAC-DRS?

CAC-DRS (Coronary Artery Calcium Data and Reporting System) is a standardized reporting framework from the Society of Cardiovascular Computed Tomography. It classifies patients by total Agatston score category (0, 1–99, 100–299, ≥300) and the number of vessels involved, and links each class to general management recommendations, similar in spirit to BI-RADS for mammography.

Coronary artery calcium (CAC) scoring converts a fast, non-contrast, ECG-gated chest CT into a quantitative marker of atherosclerotic plaque burden — but the number is only trustworthy when the acquisition, calibration, and reconstruction are standardized and verified. The Agatston score depends on a fixed 130 HU threshold, so CT number accuracy, slice thickness, tube voltage, and reconstruction settings all move the result directly. That makes calcium scoring a medical physics problem as much as a cardiology one.

Introduction

A CAC score answers a deceptively simple question: how much calcified plaque is in this patient's coronary arteries? Calcified plaque is a direct footprint of atherosclerosis, and the quantity of it predicts future cardiovascular events better than most traditional risk factors. Because calcium is dense and high-contrast against soft tissue, it can be quantified from a low-dose, non-contrast CT in a single breath-hold. ¹

The dominant metric, the Agatston score, was defined in 1990 on electron-beam CT and has survived essentially unchanged into the multidetector and photon-counting era. ¹ Its longevity is a strength — decades of outcome data are anchored to it — but also a liability: it is tied to a fixed 130 HU threshold and a specific slice thickness, so it is exquisitely sensitive to how the scan is acquired and reconstructed. Change the tube voltage, let the CT number calibration drift, or reconstruct at the wrong slice thickness, and a patient can cross a risk-category boundary without any change in their coronary arteries. ²⁷

This article walks through what the score is, how it is calculated, why standardization and reproducibility matter, the radiation dose involved, and the regulatory and accreditation context. DRPS supports calcium scoring programs as part of its CT physics testing and accreditation support services across Florida, Maryland, Virginia, Washington DC, California, and Nevada.

Topic Explanation

What is coronary artery calcium scoring?

Coronary artery calcium scoring is the quantification of calcified coronary plaque from a non-contrast, ECG-synchronized cardiac CT, reported most often as the Agatston score. The exam is prospectively triggered near end-diastole to freeze cardiac motion, covers the heart from the carina to the diaphragm, and is reconstructed at 2.5–3 mm slice thickness. A radiologist or cardiologist identifies calcified lesions in each coronary artery, and software computes the score. ¹³

Three related metrics are produced from the same data:

Agatston score — the historical standard, weighting lesion area by peak CT number. It is the metric with the deepest outcome evidence. ¹
Volume score — the calcified volume above threshold, in mm³. It is more reproducible than Agatston in some settings but shows greater interscanner variance. ⁴⁸
Calcium mass score — the calcified mass in milligrams of calcium hydroxyapatite, derived from a calibrated density measurement. It has the smallest variance across scanners and patient sizes when a density threshold is used. ²

Key terms

Hounsfield unit (HU) — the CT number scale, fixed so that water is 0 HU and air is −1000 HU. Calcium scoring uses a lesion detection threshold of 130 HU. ¹
Agatston weighting factor — an integer (1–4) assigned to each lesion based on its peak HU.
CAC-DRS — the SCCT reporting system that maps total score and vessel involvement to standardized risk classes. ³
Interscan variability — the percentage difference between two calcium scores acquired on the same patient in the same session, a direct measure of reproducibility. ⁴⁵⁶

For background on the CT number scale that underpins all of this, see our guide to CT number and HU calibration QC.

Key Technical Principles

The Agatston score definition

The Agatston score is computed lesion by lesion. A calcified lesion is any connected region of at least a minimum area with CT numbers at or above 130 HU. Each lesion is assigned a weighting factor $w$ based on its peak CT number: ¹

Peak CT number (HU)	Agatston weighting factor $w$
130–199	1
200–299	2
300–399	3
≥ 400	4

The score for a single lesion is its area multiplied by its weighting factor, and the total Agatston score is the sum over every lesion on every slice:

S_{Agatston} = i \sum A_{i} \cdot w (HU_{m a x, i})

where $A_{i}$ is the lesion area in mm² and $w (HU_{m a x, i})$ is the weighting factor for that lesion's peak CT number.

Worked example. Consider a single calcified plaque on one 3 mm slice with an area of 6 mm² and a peak CT number of 350 HU. Its weighting factor is $w = 3$ , so its contribution is:

S = 6 mm^{2} \times 3 = 18

A patient's total score sums contributions like this across all lesions and slices, which is why total scores routinely reach the hundreds or thousands.

Why the 130 HU threshold makes calibration critical

Because lesion detection and weighting both hinge on absolute CT numbers, the score is only as accurate as the scanner's HU calibration. If daily water-phantom QC shows the CT number of water drifting from 0 HU toward, say, +5 HU, calcified voxels near the 130 HU or the 200/300/400 HU boundaries shift weighting factors, and the total score changes with no change in the patient. This is the core reason a calcium scoring program needs verified CT number calibration, standardized reconstruction, and a controlled tube voltage. ²⁷

Tube voltage is especially important. The 130 HU threshold and the Agatston method were defined at 120 kVp, so 120 kVp is the standard acquisition voltage for calcium scoring. Lowering kVp raises the CT number of calcium (and iodine), which would inflate scores if the fixed threshold and weighting boundaries were applied naively. Modern approaches either hold 120 kVp or apply validated kV-specific corrections. ²⁷

The calcium mass score and hydroxyapatite calibration

The mass score addresses the scanner dependence of the Agatston and volume scores. It uses a calibration factor $c$ that relates measured CT numbers to calcium hydroxyapatite density, obtained by scanning a phantom with inserts of known hydroxyapatite concentration. The calcium mass in a lesion is approximately:

m = c \cdot V \cdot \overline{CT}

where $V$ is the calcified volume and $\overline{CT}$ is the mean CT number above threshold. The multi-institutional standardization work of McCollough and colleagues showed that using a fixed hydroxyapatite density threshold of 100 mg/cm³ — instead of a fixed 130 HU threshold — reduces interscanner variability in Agatston and mass scores, and that the mass score has the smallest variance as a function of patient size. ²

Reproducibility and interscan variability

Interscan variability quantifies how much a repeated score differs. For two scans A and B on the same patient, it is defined as: ⁴

V_{interscan} = \frac{∣ S _{A} - S _{B} ∣}{\frac{1}{2} ( S _{A} + S _{B} )} \times 100%

Reported mean interscan variability for well-controlled ECG-gated protocols is on the order of 10–15% for the Agatston, volume, and mass scores. In a low-dose prospectively triggered study, Horiguchi and colleagues found roughly 13% interscan variability for the Agatston score while keeping the effective dose near 0.9 mSv. ⁶ The CARE-2 dual-scan analysis reported about 11.8% Agatston variability overall. ⁴

Crucially, variability is not uniform across score magnitudes. It is small for large scores and very large for small ones. In CARE-2, Agatston variability was about 8% for scores above 400 but exceeded 60% for scores of 1–30. ⁴ Ungated low-dose acquisitions — for example, CAC read off a lung-cancer screening CT — show far higher variability still, with mean Agatston interscan variability reported near 70%, which precludes using those scans to track change in an individual. ⁵

Newer detector technology is narrowing this. A 2026 phantom study on photon-counting detector CT found that a thin-slice, dose-reduced 120 kVp protocol cut score variability by roughly a third compared with the standard photon-counting protocol and outperformed the proposed energy-integrating-detector protocol, pointing toward more reproducible quantification at lower dose. ⁷ The companion topic of detector physics is covered in our overview of photon-counting CT image quality.

CAC-DRS reporting

The Society of Cardiovascular Computed Tomography's CAC-DRS framework standardizes how scores are communicated. It classifies patients by total Agatston score category and by the number of coronary arteries involved (N0–N4), and attaches general management guidance to each class. ³

Total Agatston score	CAC-DRS category	Calcified plaque burden	General clinical framing
0	A0	None detected	Lowest near-term risk; supports deferring or de-escalating therapy in many patients
1–99	A1	Mild	Mild plaque; consider risk-factor modification
100–299	A2	Moderate	Moderate plaque; risk-factor modification and often preventive pharmacotherapy
≥ 300	A3	Severe	Severe plaque; intensified preventive management

CAC-DRS plays a role analogous to BI-RADS in mammography: a shared, structured vocabulary that ties a quantitative finding to a recommended action and supports quality tracking. ³

Clinical Impact

A calcium score reclassifies cardiovascular risk in a way few other single tests can. A score of 0 in an asymptomatic intermediate-risk patient is a powerful negative finding that can justify deferring statin therapy, while a high score can move a borderline patient decisively toward aggressive prevention. Because the exam is fast, non-contrast, and low-dose, it is well suited to asymptomatic risk stratification. ¹³

The physics choices ripple straight into that clinical decision. A patient sitting at an Agatston score near 100 is at a category boundary; a 10–15% reproducibility band, or a CT number calibration drift, can push them from "mild" to "moderate," changing the management conversation. This is why the standardization work matters clinically and not just academically — it is the difference between a number a clinician can act on and one they must second-guess. ²⁴

Serial scoring — using change in CAC over time to judge progression — is where reproducibility limits bite hardest. Because small scores carry large variability, a change has to exceed the interscan noise before it can be called real. Programs that image the same patient over years should hold the protocol, scanner, and reconstruction as constant as possible so that a change in the number reflects the patient, not the technique. ⁴⁵

Practical Optimization Tips

1. Lock the acquisition protocol to the standardized method

Use 120 kVp, prospective ECG triggering near end-diastole, and reconstruct at 2.5–3 mm slice thickness with a standard (non-sharp) kernel. These are the conditions under which the Agatston method and its outcome data were established. Deviating from them without validated corrections breaks comparability. ¹²³

2. Verify CT number and calcium calibration

Confirm that daily water-phantom QC keeps water near 0 HU, and periodically scan a calcium hydroxyapatite calibration phantom to establish the density-to-CT-number relationship for the mass score. Calibration drift is a silent way to shift scores across risk categories. ²

3. Standardize patient-size handling

Image noise rises with patient size and inflates low-density score variability. Use size-appropriate tube current to hold a target noise level, as recommended in the multi-institutional standard, rather than a single fixed technique for every patient. ²

4. Keep dose as low as reproducibility allows

Calcium scoring is inherently low-dose, but dose and noise trade against reproducibility. Prospective ECG triggering, appropriate tube current, and modern reconstruction can hold the effective dose near or below about 1 mSv while preserving acceptable score variability. Track dose with the same rigor used elsewhere in CT — see our guide to CTDIvol and DLP dose metrics. ⁶⁷

5. Hold everything constant for serial exams

When a patient is scanned repeatedly over years, keep the scanner, protocol, kVp, slice thickness, and reconstruction identical. Changing any of them introduces a technique-driven score shift that can masquerade as disease progression or regression. ⁴⁵

Common pitfalls to avoid

Letting CT number calibration drift. A few HU of offset moves lesions across weighting boundaries.
Reconstructing at the wrong slice thickness. Thinner or thicker slices than the standard 3 mm change the score.
Using ungated or lung-screening CT for serial tracking. Interscan variability is far too high for individual follow-up.
Ignoring patient size. One-size-fits-all technique produces size-dependent scores.
Changing scanners mid-program without cross-calibration. Volume and Agatston scores carry meaningful interscanner variance. ²⁵

Regulatory Considerations

Calcium scoring is performed on a CT scanner, so it sits under the regulatory framework for radiation-producing machines, not byproduct material. CT systems must meet the FDA electronic-product performance standard at 21 CFR 1020.33, and clinical use is regulated by state radiation-control programs. ⁹

Key frameworks:

FDA 21 CFR 1020.33 — the federal performance standard for CT equipment, including dose information requirements. ⁹
State radiation-control rules. X-ray machines are registered and inspected by the state. In Florida, CT operation falls under Florida Administrative Code Chapter 64E-5, Part V, for radiation machines; DRPS also serves Maryland, Virginia, Washington DC, California, Nevada, and other states where parallel state programs apply. Always confirm requirements with the authority having jurisdiction.
ACR CT Accreditation and the ACR–NASCI–SPR practice parameter for cardiac CT. Facilities pursuing accreditation must demonstrate qualified-physicist involvement, dose monitoring, and image-quality performance. Cardiac and calcium scoring protocols are part of that review. ¹⁰
NCRP Report No. 147 — the structural shielding design methodology for medical X-ray imaging facilities, relevant when a CT suite is planned or renovated. ¹¹

A qualified medical physicist's annual CT performance evaluation, CT number calibration verification, and dose review are what make a calcium scoring program defensible under accreditation and state inspection. For the broader accreditation picture, see our overview of ACR accreditation physics requirements, and connect the program to medical physics consulting when protocols change.

Frequently Asked Questions (FAQs)

What is a coronary artery calcium (CAC) score?

A CAC score is a quantitative measure of calcified atherosclerotic plaque in the coronary arteries, derived from a non-contrast ECG-gated chest CT. The most common metric is the Agatston score, which weights each calcified lesion by its area and its peak CT number and sums the result across all coronary arteries.

What is a normal or high calcium score?

An Agatston score of 0 indicates no detectable calcified plaque and low near-term risk. Scores of 1–99 indicate mild plaque, 100–299 moderate, and 300 or above severe plaque burden. The number is interpreted together with the patient's age, sex, and risk factors.

How much radiation does a calcium scoring CT deliver?

A dedicated calcium scoring CT is a low-dose exam, with typical effective doses on the order of about 1 mSv using modern prospective ECG-triggered protocols. The exact value depends on scanner, patient size, and technique, and it is far lower than a contrast-enhanced coronary CT angiogram.

Why does the CT protocol matter so much for calcium scoring?

The Agatston score depends on a fixed 130 HU threshold, so drift in CT number calibration or changes in slice thickness, tube voltage, or reconstruction can shift the score across a risk category. Standardized acquisition — 120 kVp, 2.5–3 mm slices, ECG gating, and verified HU calibration — keeps scores comparable over time and between scanners.

How reproducible is the calcium score?

Repeat scores typically vary by roughly 10–15% for scores above about 100, but variability is much larger for very low scores and for ungated acquisitions. This interscan variability limits how confidently small changes can be called real progression.

What is CAC-DRS?

CAC-DRS is a standardized reporting framework from the Society of Cardiovascular Computed Tomography. It classifies patients by total Agatston score category and the number of vessels involved and links each class to general management recommendations, similar in spirit to BI-RADS for mammography.

Should a medical physicist be involved in a calcium scoring program?

Yes. A qualified medical physicist verifies CT number and calcium calibration, confirms the acquisition protocol matches the standardized method, tracks radiation dose, and supports ACR CT accreditation — keeping scores reproducible and defensible.

Key Takeaways

The Agatston score is threshold-dependent. Lesion detection at 130 HU and weighting at 200/300/400 HU boundaries make absolute CT number accuracy essential. ¹
Standardization is the whole game. 120 kVp, 2.5–3 mm slices, ECG gating, verified calibration, and size-appropriate technique are what make scores comparable. ²³
Mass scoring is the most robust metric. A hydroxyapatite density threshold reduces scanner and patient-size dependence. ²
Reproducibility is score-dependent. Roughly 10–15% interscan variability for larger scores, but much higher for small scores and ungated scans. ⁴⁵⁶
CAC-DRS structures the report. Total score category plus vessel involvement, tied to management guidance. ³
It is a low-dose exam that still deserves dose tracking. Modern protocols hold effective dose near 1 mSv. ⁶⁷

Conclusion

Coronary artery calcium scoring is one of the most clinically valuable low-dose CT exams in preventive cardiology, but its value rests entirely on the reproducibility of a threshold-based number. The Agatston score's fixed 130 HU dependence means that CT number calibration, tube voltage, slice thickness, reconstruction, and patient-size handling are not background details — they are the determinants of whether a patient is called mild, moderate, or severe. A calcium scoring program that standardizes acquisition, verifies calibration, tracks dose, and holds technique constant for serial exams produces scores clinicians can act on with confidence. That is a medical physics responsibility, and it is what turns a quantitative image into a durable clinical decision. ¹²³

How DRPS Can Help

Diagnostic Radiation Physics Services helps CT and cardiac imaging facilities keep calcium scoring reproducible and defensible. This includes CT physics testing, CT number and calcium calibration verification, calcium scoring protocol review, radiation dose assessment, and ACR accreditation support prepared by board-certified medical physicists.

DRPS supports facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware.

A strong calcium scoring program is not just about passing accreditation. It is about making sure the number a clinician reads reflects the patient's arteries, not the day's technique.

Related Resources

References

Agatston AS, Janowitz WR, Hildner FJ, et al. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol. 1990;15(4):827-832. doi:10.1016/0735-1097(90)90282-T. doi.org
McCollough CH, Ulzheimer S, Halliburton SS, et al. Coronary artery calcium: a multi-institutional, multimanufacturer international standard for quantification at cardiac CT. Radiology. 2007;243(2):527-538. doi:10.1148/radiol.2432050808. doi.org
Hecht HS, Blaha MJ, Kazerooni EA, et al. CAC-DRS: Coronary Artery Calcium Data and Reporting System. An expert consensus document of the Society of Cardiovascular Computed Tomography (SCCT). J Cardiovasc Comput Tomogr. 2018;12(3):185-191. doi:10.1016/j.jcct.2018.03.008. doi.org
Budoff MJ, Kessler P, Gao YL, et al. The interscan variation of CT coronary artery calcification score: analysis of the Calcium Acetate Renagel Comparison (CARE)-2 study. Acad Radiol. 2008;15(1):58-61. doi:10.1016/j.acra.2007.08.011. doi.org
Jacobs PC, Isgum I, Gondrie MJ, et al. Coronary artery calcification scoring in low-dose ungated CT screening for lung cancer: interscan agreement. AJR Am J Roentgenol. 2010;194(5):1244-1249. doi:10.2214/AJR.09.3047. doi.org
Horiguchi J, Matsuura N, Yamamoto H, et al. Coronary artery calcium scoring on low-dose prospective electrocardiographically-triggered 64-slice CT. Acad Radiol. 2009;16(2):187-193. doi:10.1016/j.acra.2008.05.017. doi.org
Fink N, Koetzier LR, Zsarnoczay E, et al. Coronary artery calcium scoring: expanding the new standard by photon-counting detector CT — Part I. Eur Radiol. 2026;36(7):5546-5556. doi:10.1007/s00330-026-12355-4. doi.org
Horiguchi J, Yamamoto H, Hirai N, et al. Variability of repeated coronary artery calcium measurements on low-dose ECG-gated 16-MDCT. AJR Am J Roentgenol. 2006;187(1):W1-W6. doi:10.2214/AJR.05.0052. doi.org
U.S. Food and Drug Administration. 21 CFR 1020.33 — Computed Tomography (CT) Equipment. ecfr.gov
American College of Radiology. ACR Practice Parameters and Technical Standards (Cardiac CT; CT Accreditation). acr.org
National Council on Radiation Protection and Measurements. Structural Shielding Design for Medical X-Ray Imaging Facilities. NCRP Report No. 147. Bethesda, MD: NCRP; 2004. aapm.org