Mean Glandular Dose in Mammography

Mean glandular dose (MGD) is the accepted measure of radiation dose in mammography because the fibroglandular tissue of the breast is the radiosensitive target. MGD cannot be measured directly inside the compressed breast, so a medical physicist estimates it by multiplying a measured incident air kerma by published Monte Carlo conversion factors that depend on breast thickness, glandularity, and beam quality.¹⁴⁸

Under the Mammography Quality Standards Act (MQSA), the average glandular dose from a single view of a standard-breast phantom must not exceed 3.0 mGy per exposure.² This guide explains what MGD is, the Dance and Boone calculation methods, how tomosynthesis changes the estimate, and how the dose is verified during the annual survey.

Introduction

Mammography is unusual among diagnostic imaging examinations: it images one of the most radiosensitive tissues in the body, it is performed repeatedly on healthy, asymptomatic women as a screening test, and it operates at the low end of the diagnostic energy spectrum where soft-tissue contrast and absorbed dose are both highly sensitive to beam quality. For all three reasons, dose must be quantified carefully and reported in a way that reflects biological risk rather than a convenient surface measurement.¹¹¹

That is why the field standardized on mean glandular dose decades ago. Skin dose and entrance exposure are easy to measure but poorly correlated with risk, because the glandular tissue at depth receives a fraction of the surface dose that depends strongly on breast thickness and X-ray penetration. MGD reframes the dose question around the tissue that actually carries the radiation-induced cancer risk.¹¹

This article walks through the definition of MGD, the conversion-factor formalisms that turn a measurable air kerma into an estimated glandular dose, the regulatory dose limit, the differences introduced by digital breast tomosynthesis (DBT), and the practical steps a physicist takes to verify dose during the annual MQSA evaluation. DRPS provides this analysis as part of its mammography physics and MQSA services across Florida, Maryland, Virginia, Washington DC, California, and Nevada.

Topic Explanation

What is mean glandular dose?

Mean glandular dose is the average absorbed dose to the glandular tissue of the breast produced by a mammographic exposure. It is expressed in milligray (mGy) and is the dose metric required by accreditation bodies and regulators for both screening and diagnostic mammography.¹³

The key conceptual points are:

• The target tissue is the fibroglandular tissue, not skin or adipose tissue. Glandular tissue is where breast cancers arise and where radiation risk is concentrated.
• MGD is a calculated, not directly measured, quantity. No instrument can sit inside a compressed breast and read glandular dose.
• The conversion from a measurable radiation quantity to MGD depends on breast thickness, breast composition (glandularity), and beam quality (kVp, target, filter, and half-value layer).

For background on how dose metrics are defined and reported in other modalities, see our explainer on CTDIvol and DLP dose metrics, and for how mammographic image quality is verified alongside dose, see Mammography Quality Control under MQSA.

Why not just measure entrance dose?

Entrance skin dose and incident air kerma are easy to measure with an ion chamber or solid-state detector placed at the breast support surface. But the relationship between surface dose and glandular dose is not fixed. A thick, fatty breast attenuates the beam very differently from a thin, dense breast, and the low-energy mammographic spectrum is heavily filtered as it passes through tissue. Two patients with the same entrance air kerma can receive substantially different glandular doses.¹¹¹

The dosimetry community resolved this by measuring the quantity that is reproducible (incident air kerma) and converting it to the quantity that matters (glandular dose) using Monte Carlo-derived coefficients. This is the foundation of every modern breast dosimetry protocol.⁴⁵⁶

Key Technical Principles

The Dance formalism

The most widely used breast dosimetry formalism in Europe, the United Kingdom, and IAEA protocols was developed by Dance and colleagues. The mean glandular dose is estimated as:⁴⁵⁶

$$\text{MGD}=K\cdot g\cdot c\cdot s$$

where:

• $K$ is the incident air kerma at the upper surface of the breast (measured with no breast or phantom present, at the position of the breast surface).
• $g$ is the conversion factor from air kerma to glandular dose for a breast of 50% glandularity, tabulated as a function of breast thickness and half-value layer (HVL). Introduced in Dance's 1990 Monte Carlo work.⁴
• $c$ is the glandularity correction, which accounts for breast composition differing from the 50% reference, as a function of thickness and HVL.⁵
• $s$ is the spectrum factor, which corrects for the specific target/filter combination (for example Mo/Mo, Mo/Rh, or W/Rh).⁵⁶

For digital breast tomosynthesis, the formalism is extended with a geometry term:⁷

$${\text{MGD}}_{\text{DBT}}=K\cdot g\cdot c\cdot s\cdot T$$

where $T$ is the tomosynthesis factor accounting for the angular acquisition geometry of the sweeping X-ray tube.⁷

The Boone formalism

In the United States, the normalized glandular dose (DgN) approach developed by Boone is also common. Here the glandular dose $D_g$ is:⁸

$$D_g=X\cdot \text{DgN}$$

where $X$ is the measured entrance exposure or incident air kerma, and $\text{DgN}$ is the normalized glandular dose coefficient — the glandular dose per unit entrance air kerma. Boone published computer-fit DgN values for arbitrary X-ray spectra as a function of beam energy, breast thickness, and glandular fraction, making the coefficients usable for any clinical spectrum.⁸

The two formalisms are conceptually identical: both multiply a measurable radiation quantity by a Monte Carlo-derived coefficient that encodes how much of the incident beam is absorbed in glandular tissue. The 2024 joint AAPM Task Group 282 / EFOMP report consolidated these approaches into a single universal breast dosimetry method covering standard mammography, tomosynthesis, and contrast-enhanced mammography for both CC and MLO views.¹⁰

Comparing the two formalisms

Feature	Dance formalism	Boone (DgN) formalism
Core equation	$\text{MGD}=Kgcs$	$D_g=X\cdot \text{DgN}$
Measured quantity	Incident air kerma $K$	Entrance exposure / air kerma $X$
Reference breast	50% glandular, thickness-dependent	Variable glandular fraction
Glandularity handling	Separate $c$ factor	Built into DgN coefficient
Spectrum handling	Separate $s$ factor	Built into DgN coefficient
Tomosynthesis	$T$ factor extension (2011)	Geometry-specific coefficients
Primary regions of use	UK, Europe, IAEA	United States
Unifying reference	AAPM TG-282 / EFOMP (2024)	AAPM TG-282 / EFOMP (2024)

Worked MGD example (Dance method)

Consider a standard-breast phantom exposure under clinical automatic exposure control with a tungsten/rhodium (W/Rh) spectrum:

Assumptions:

• Incident air kerma at the breast surface: $K=8.0\text{mGy}$.
• $g$-factor for a 4.2 cm breast at the measured HVL: $g=0.20$.⁴
• Glandularity correction for the phantom: $c=1.05$.⁵
• Spectrum factor for W/Rh: $s=1.01$.⁶

The estimated mean glandular dose is:

$$\text{MGD}=8.0\times 0.20\times 1.05\times 1.01\approx 1.70\text{mGy}$$

This result, about 1.7 mGy, is well below the 3.0 mGy regulatory limit and is consistent with typical digital mammography doses reported in large clinical series.²¹²

If the same acquisition were a tomosynthesis sweep with $T=1.05$, the estimate would rise to approximately:

$${\text{MGD}}_{\text{DBT}}=1.70\times 1.05\approx 1.79\text{mGy}$$

still comfortably below the limit, illustrating why a single DBT acquisition is only modestly higher in dose than a 2D view.⁷¹²

Clinical Impact

Screening at scale magnifies small differences

Because mammography is a population screening test, small per-exposure dose differences are repeated across millions of examinations and across a woman's screening lifetime. A standard screening study includes CC and MLO views of each breast, and many protocols now combine 2D and tomosynthesis acquisitions or use synthesized 2D images to control dose. Understanding MGD per acquisition is what allows a program to add tomosynthesis without an unacceptable dose increase.⁷¹²

In a large paired study of same-compression acquisitions, the mean average glandular dose was approximately 1.74 mGy for digital mammography and 2.10 mGy for tomosynthesis, with a DBT/DM ratio near 1.24 overall. Importantly, the dose penalty for tomosynthesis was much smaller for dense breasts than for fatty breasts.¹² This is clinically reassuring, because dense breasts are exactly where tomosynthesis adds the most diagnostic value.

Dose, density, and risk communication

MGD also underpins how facilities discuss radiation risk with patients, especially now that the 2023 MQSA Final Rule requires breast density reporting in mammography results.² When a patient asks whether tomosynthesis "doubles the dose" or whether annual screening is safe, the answer is grounded in MGD values that are typically a small fraction of natural background dose accumulated over a year. A physicist-verified MGD is the technical foundation for that reassurance.

Image quality cannot be separated from dose

Lowering dose is only beneficial if image quality is preserved. Mammography must resolve microcalcifications a few hundred micrometers across and detect low-contrast masses against a structured glandular background. The automatic exposure control (AEC) selects kVp, target, filter, and mAs to balance dose against noise and contrast. A defensible dose evaluation always pairs the MGD measurement with phantom image quality scoring and signal-to-noise analysis, so that a "low dose" result is not actually an underexposed, diagnostically inadequate image.³⁹

Practical Optimization Tips

A practical MGD evaluation during the annual survey generally follows these steps.

1. Reproduce clinical conditions

Expose the FDA-accepted phantom under the same AEC mode the clinic uses for patients. The goal is to characterize the dose patients actually receive, not an idealized manual technique.

2. Record the technique factors

Capture the kVp, target, filter, compression thickness, and mAs that the AEC selects. These determine which conversion factors apply.

3. Measure incident air kerma and beam quality

Use a calibrated mammography dosimeter to measure air kerma at the breast surface position, and measure the half-value layer to characterize beam quality, which drives the $g$ and $c$ factors.⁴⁵

4. Apply the correct conversion factors

Select $g$, $c$, and $s$ (or the appropriate DgN coefficient) for the measured thickness, glandularity, and spectrum. For tomosynthesis, include the $T$ factor.⁶⁷

5. Compare against the limit and the baseline

Confirm MGD is below 3.0 mGy and consistent with the unit's established baseline. A sudden rise in dose at constant image quality can indicate AEC drift, detector degradation, or filtration changes.

Common pitfalls to avoid

• Reporting entrance dose as if it were glandular dose. They are not interchangeable; the conversion is essential.
• Using the wrong target/filter spectrum factor. Modern units commonly switch automatically between Mo/Mo, Mo/Rh, and W/Rh; the $s$ factor must match.
• Ignoring glandularity. A 50% reference is convenient but the $c$ factor exists precisely because real breasts and phantoms differ.
• Evaluating dose without image quality. A low MGD that fails phantom scoring is not a success.
• Forgetting the tomosynthesis geometry factor. Omitting $T$ underestimates DBT dose.
• Treating the 3.0 mGy limit as a target. It is a regulatory ceiling for a standard-breast phantom, not an operating goal; clinical doses should be well below it.

Regulatory Considerations

Mammography dose is one of the most tightly regulated quantities in diagnostic imaging. In the United States, mammography is governed by the Mammography Quality Standards Act and its implementing regulations at 21 CFR Part 900.²

Key requirements:

• 21 CFR 900.12(e)(5)(vi) sets the dose limit: the average glandular dose from a single craniocaudal view of an FDA-accepted phantom simulating a standard breast must not exceed 3.0 mGy (0.3 rad) per exposure. The standard breast is a 4.2 cm thick compressed breast of 50% glandular and 50% adipose tissue.²
• Annual medical physicist evaluation is required under MQSA, including dose measurement, beam quality, AEC performance, and phantom image quality.²³
• The 2023 MQSA Final Rule (88 FR 15126), enforced beginning September 10, 2024, modernized several requirements, most notably mandatory breast density reporting. The 3.0 mGy dose limit itself was unchanged.²
• Accreditation under the ACR Digital Mammography QC Manual (2018 edition) establishes the operational QC program, including the technologist and physicist tests that support dose and image quality compliance.³

Because X-ray mammography units are regulated by the FDA together with state radiation-control programs, facilities must satisfy both federal MQSA requirements and any additional state rules. Of the states DRPS serves, Florida, Maryland, Virginia, California, and Nevada are NRC Agreement States with their own radiation-control programs for X-ray equipment, while Washington, DC is regulated directly at the federal level. For state-specific context, see Florida Radiation Safety Requirements for Imaging Centers, and to understand how dose and image quality fit into accreditation, see ACR Accreditation Physics Requirements.

Facilities preparing for accreditation or an MQSA inspection should connect the dose evaluation to documented QC, physicist reports, and corrective-action records as part of accreditation support and medical physics consulting.

Frequently Asked Questions (FAQs)

What is mean glandular dose (MGD) in mammography?

Mean glandular dose is the average absorbed dose to the glandular tissue of the breast from a mammographic exposure. It is used instead of skin or entrance dose because the fibroglandular tissue is the radiosensitive structure at risk. MGD cannot be measured directly, so it is estimated by multiplying a measured incident air kerma by published conversion factors that depend on breast thickness, glandularity, and X-ray beam quality.

What is the MQSA dose limit for mammography?

Under 21 CFR 900.12(e)(5)(vi), the average glandular dose from a single craniocaudal view of an FDA-accepted phantom that simulates a standard breast must not exceed 3.0 mGy (0.3 rad) per exposure. The standard breast is defined as a 4.2 cm thick compressed breast composed of 50% glandular and 50% adipose tissue. This is a regulatory action limit, not a target; typical clinical doses are well below it.

How is mean glandular dose calculated?

Two equivalent formalisms are used. The Dance method multiplies the incident air kerma K by factors g (thickness and beam quality), c (glandularity), and s (target/filter spectrum): MGD = K g c s. The Boone method multiplies the entrance exposure or air kerma X by a normalized glandular dose coefficient DgN: MGD = X times DgN. Both rely on Monte Carlo-derived coefficients tabulated against breast thickness, glandular fraction, and beam quality.

Does digital breast tomosynthesis deliver more dose than 2D mammography?

A single tomosynthesis acquisition typically delivers a slightly higher mean glandular dose than a single 2D digital view, but the difference is modest and depends on breast composition. In a large paired study the mean was about 1.74 mGy for digital mammography and about 2.10 mGy for tomosynthesis per acquisition, with a smaller dose penalty for dense breasts than for fatty breasts. The MGD for tomosynthesis is estimated by adding a geometry T-factor to the Dance equation.

Who measures mean glandular dose at my facility?

A qualified medical physicist measures MGD during the annual mammography equipment evaluation required under MQSA. The physicist exposes an FDA-accepted phantom under clinical automatic exposure control, measures the incident air kerma and beam quality, applies the appropriate conversion factors, and confirms the result is below the 3.0 mGy limit and consistent with image quality requirements.

Why is incident air kerma used instead of measuring dose in the breast?

Absorbed dose to glandular tissue inside a compressed breast cannot be measured with an external instrument. Incident air kerma at the breast surface is measurable and reproducible, so dosimetry protocols measure that quantity and convert it to MGD using validated Monte Carlo coefficients. This keeps the measurement practical while still reporting the dose to the tissue that matters for risk.

Can lowering dose hurt image quality in mammography?

Yes. Mammography must balance dose against the ability to detect small, low-contrast lesions and microcalcifications. Reducing dose too far raises image noise and can mask findings. The physicist evaluates dose and image quality together, confirming that the automatic exposure control selects technique factors that keep MGD low while meeting phantom image quality and signal-to-noise criteria.

Key Takeaways

• MGD measures dose to the tissue that matters. Glandular tissue carries the radiation risk, so dose is reported as mean glandular dose rather than skin or entrance dose.
• MGD is calculated, not measured directly. A measurable incident air kerma is converted to glandular dose using Monte Carlo coefficients that depend on thickness, glandularity, and beam quality.
• Two equivalent formalisms exist. The Dance method ($Kgcs$) and the Boone DgN method ($X\cdot \text{DgN}$) are unified in the 2024 AAPM TG-282 / EFOMP universal method.
• The MQSA limit is 3.0 mGy per view for a standard-breast phantom, and typical clinical doses are well below it.
• Tomosynthesis adds only a modest dose increase per acquisition, estimated with a geometry $T$ factor, and the increase is smallest for dense breasts.
• Dose and image quality must be evaluated together so that low dose never comes at the cost of diagnostic performance.

Conclusion

Mean glandular dose is the bridge between a practical, measurable radiation quantity and the biologically meaningful dose to the breast's radiosensitive tissue. By measuring incident air kerma and beam quality and applying validated conversion factors, a medical physicist can report a dose that reflects real risk, confirm compliance with the 3.0 mGy MQSA limit, and verify that dose and image quality remain in balance.

As tomosynthesis, contrast-enhanced mammography, and synthesized 2D imaging become routine, the dosimetry must keep pace. The 2024 AAPM TG-282 / EFOMP universal method gives the field a single, consistent framework for all of these acquisitions. Facilities that treat MGD as a rigorously measured, documented quantity, rather than a box to check, are best positioned to keep dose low, image quality high, and accreditation defensible.

How DRPS Can Help

Diagnostic Radiation Physics Services helps mammography facilities measure, document, and optimize mean glandular dose as part of the annual MQSA equipment evaluation. This includes incident air kerma and HVL measurement, AEC performance assessment, conversion-factor selection for 2D and tomosynthesis acquisitions, phantom image quality scoring, baseline trending, and corrective-action support, all documented for accreditation and inspection.

DRPS supports facilities across our service locations, including Florida, Maryland, Virginia, Washington DC, California, Nevada, New York, Pennsylvania, New Jersey, and Delaware. To discuss a mammography physics survey, contact our team.

Related Resources

• Mammography Quality Control under MQSA
• Digital Breast Tomosynthesis QC
• ACR Accreditation Physics Requirements
• CTDIvol and DLP Dose Metrics
• Fetal Dose in Medical Imaging
• Mammography physics and MQSA services
• Accreditation support
• Medical physics consulting

References

1. Dance DR, Sechopoulos I. Dosimetry in x-ray-based breast imaging. Physics in Medicine and Biology. 2016;61(19):R271-R304. doi:10.1088/0031-9155/61/19/R271. PubMed
2. U.S. Food and Drug Administration. 21 CFR Part 900 — Mammography Quality Standards Act; § 900.12 Quality standards; Mammography Quality Standards Act Final Rule, 88 FR 15126 (2023). ecfr.gov
3. American College of Radiology. ACR Digital Mammography Quality Control Manual. 2018 edition. acr.org
4. Dance DR. Monte Carlo calculation of conversion factors for the estimation of mean glandular breast dose. Physics in Medicine and Biology. 1990;35(9):1211-1219. doi:10.1088/0031-9155/35/9/002. PubMed
5. Dance DR, Skinner CL, Young KC, Beckett JR, Kotre CJ. Additional factors for the estimation of mean glandular breast dose using the UK mammography dosimetry protocol. Physics in Medicine and Biology. 2000;45(11):3225-3240. doi:10.1088/0031-9155/45/11/308. PubMed
6. Dance DR, Young KC, van Engen RE. Further factors for the estimation of mean glandular dose using the United Kingdom, European and IAEA breast dosimetry protocols. Physics in Medicine and Biology. 2009;54(14):4361-4372. doi:10.1088/0031-9155/54/14/002. PubMed
7. Dance DR, Young KC, van Engen RE. Estimation of mean glandular dose for breast tomosynthesis: factors for use with the UK, European and IAEA breast dosimetry protocols. Physics in Medicine and Biology. 2011;56(2):453-471. doi:10.1088/0031-9155/56/2/011. PubMed
8. Boone JM. Normalized glandular dose (DgN) coefficients for arbitrary X-ray spectra in mammography: computer-fit values of Monte Carlo derived data. Medical Physics. 2002;29(5):869-875. doi:10.1118/1.1472499. PubMed
9. Sechopoulos I, Sabol JM, Berglund J, et al. Radiation dosimetry in digital breast tomosynthesis: report of AAPM Tomosynthesis Subcommittee Task Group 223. Medical Physics. 2014;41(9):091501. doi:10.1118/1.4892600. PubMed
10. Sechopoulos I, Dance DR, Boone JM, et al. Joint AAPM Task Group 282/EFOMP Working Group Report: Breast dosimetry for standard and contrast-enhanced mammography and breast tomosynthesis. Medical Physics. 2024;51(2):712-739. doi:10.1002/mp.16842. PubMed
11. Østerås BH, Skaane P, Gullien R, Martinsen ACT. Average glandular dose in paired digital mammography and digital breast tomosynthesis acquisitions in a population based screening program. Physics in Medicine and Biology. 2018;63(3):035006. doi:10.1088/1361-6560/aaa614. PubMed
12. National Council on Radiation Protection and Measurements. NCRP Report No. 149: A Guide to Mammography and Other Breast Imaging Procedures. 2004. ncrponline.org