Mammography remains the cornerstone of breast cancer screening worldwide, saving lives through early detection of malignancies before they become palpable or symptomatic. This comprehensive guide explains how modern breast imaging works, the critical differences between 2D digital mammography and 3D digital breast tomosynthesis, what dense breast tissue means for your screening outcomes, and why a callback request is far more common—and less alarming—than most patients initially believe.

Breast cancer is the most commonly diagnosed cancer among women globally, with approximately 2.3 million new cases and 670,000 deaths reported annually according to 2022 WHO Global Cancer Observatory data^[1]. When detected at Stage 0 or Stage 1, the five-year relative survival rate exceeds 99%, underscoring the life-saving value of routine mammography screening^[2]. This article is designed for patients, caregivers, radiographers, and hospital administrators seeking authoritative, evidence-based guidance on every aspect of the mammography journey—from preparation and compression to understanding callbacks and dense-breast notification letters.

Clinical background and pathophysiology

Breast tissue is composed of three primary elements: glandular tissue (lobules and ducts), fibrous connective tissue (stroma), and adipose (fat) tissue. The relative proportion of these components determines breast density, which is a major factor in both mammographic sensitivity and independent breast cancer risk. On a mammogram, fat appears dark (radiolucent), while glandular and fibrous tissue appears white (radiopaque). Because most breast cancers also appear white, dense breast tissue can mask tumours—a phenomenon radiologists describe as masking effect or confluence.

Relevant anatomy and breast density categories

The ACR Breast Imaging Reporting and Data System (BI-RADS) classifies breast density into four categories based on the proportion of dense fibroglandular tissue visible on a mammogram^[3]:

• Category A (Almost entirely fatty): Fatty tissue comprises more than 75% of the breast. Cancers are easily visible against the dark background. Found in approximately 10% of women.
• Category B (Scattered fibroglandular densities): Some scattered areas of dense tissue, but the majority is fatty. Found in approximately 40% of women.
• Category C (Heterogeneously dense): Dense tissue comprises 51–75% of the breast. May obscure small masses. Found in approximately 40% of women.
• Category D (Extremely dense): Dense tissue comprises more than 75% of the breast. Significantly lowers mammographic sensitivity. Found in approximately 10% of women.

Women with heterogeneously dense or extremely dense breasts (Categories C and D combined, roughly 50% of the female population) face a four- to six-fold increased risk of breast cancer compared with women with almost entirely fatty breasts, independent of other risk factors^[4].

Epidemiology and risk factors

Breast cancer risk increases with age, with the majority of cases diagnosed in women over 50. However, approximately 11% of all new breast cancers occur in women under 45^[5]. Key risk factors include:

• Genetic mutations: BRCA1 and BRCA2 carriers face a lifetime risk of 45–65%.
• Family history: First-degree relative with breast cancer approximately doubles risk.
• Reproductive history: Early menarche, late menopause, nulliparity, or first pregnancy after age 30.
• Hormone replacement therapy: Combined oestrogen-progestin therapy increases risk, particularly during active use.
• Prior chest radiation: Especially therapeutic radiation for Hodgkin lymphoma before age 30.
• Alcohol consumption and obesity: Both are modifiable risk factors with dose-dependent relationships.

Imaging protocol and technique

Modern breast imaging relies on two principal mammographic techniques: 2D full-field digital mammography (FFDM) and 3D digital breast tomosynthesis (DBT). Both use low-energy X-rays transmitted through compressed breast tissue, but 3D tomosynthesis acquires multiple projection images across a limited arc (typically 15–50 degrees) to reconstruct thin slice images that reduce tissue overlap.

2D vs 3D digital tomosynthesis

Standard 2D digital mammography captures two images per breast: a craniocaudal (CC) view and a mediolateral oblique (MLO) view. While effective for fatty or mildly dense breasts, 2D imaging suffers from structure noise—the superimposition of normal fibroglandular tissue that can mimic or obscure true lesions.

3D tomosynthesis addresses this limitation by acquiring a series of low-dose projection images (typically 9–15 exposures) as the X-ray tube arcs across the breast. A computer reconstructs these into 1-mm slice images, effectively peeling away overlapping tissue layers. Meta-analyses demonstrate that combining 2D + 3D tomosynthesis (synthetic 2D) improves invasive cancer detection by 40–53% and reduces false-positive callbacks by 15–37% compared with 2D alone^[6].

Patient preparation and positioning

Proper preparation ensures optimal image quality and minimises the need for repeat exposures. Patients should follow these evidence-based steps:

1. Schedule timing: For premenopausal women, schedule during the first half of the menstrual cycle (days 7–14) when breasts are least tender and least dense.
2. Skin products: Do not apply deodorant, antiperspirant, powder, lotion, or perfume to the underarms or breast area on the day of the exam. Aluminium particles and talc create dense white artefacts that mimic calcifications.
3. Clothing: Wear a two-piece outfit for convenience. The facility will provide a gown, but you may keep bottoms on.
4. Prior imaging: Bring prior mammogram images and reports, or ensure they are available through your imaging network. Comparison with previous studies is essential for detecting subtle interval changes.
5. Implants: Inform the technologist if you have breast implants. Special implant-displaced (Eklund) views are required to visualise tissue beyond the implant margins.
6. Pregnancy: Inform staff if you are or may be pregnant. Mammography is generally avoided in pregnancy unless absolutely necessary; ultrasound is the preferred alternative.

Acquisition parameters and radiation dose

A standard two-view screening mammogram (CC and MLO per breast) delivers an average glandular dose of approximately 0.4 mSv per breast, or 0.8 mSv total for a bilateral study^[7]. For context, this is equivalent to approximately 7 weeks of natural background radiation in the United States. Adding 3D tomosynthesis increases total dose by roughly 0.2–0.3 mSv per breast when performed in addition to 2D, though synthetic 2D protocols eliminate this incremental exposure.

Compression is applied for 10–20 seconds per view. While uncomfortable, compression is non-negotiable for three reasons: it thins the breast to reduce geometric blur and scatter radiation, it separates overlapping tissue planes, and it holds the breast stationary to prevent motion artefact. Modern compression paddles use flexible, curved surfaces that distribute pressure more evenly than flat paddles, improving patient comfort without sacrificing image quality.

Image interpretation and diagnostic criteria

Mammogram interpretation requires systematic evaluation of breast parenchyma, skin, nipple, axillary lymph nodes, and visible chest wall structures. Radiologists use the ACR BI-RADS lexicon to describe findings with standardised terminology that ensures consistent communication across facilities and countries^[3].

Normal imaging appearance

A normal mammogram demonstrates symmetric glandular tissue distribution, smooth skin contours without thickening or retraction, a central nipple shadow, and a retromammary fat layer separating breast tissue from the pectoral muscle on MLO views. The axillary tail (tail of Spence) may contain normal lymph nodes measuring up to 2 cm with a characteristic fatty hilum. Vascular calcifications appear as parallel linear or tram-track densities and are benign age-related findings.

Pathological findings and the BI-RADS assessment

Radiologists classify every mammogram into one of seven BI-RADS assessment categories that dictate management:

• BI-RADS 0 (Incomplete): Additional imaging or comparison with prior studies is needed before a final assessment can be assigned. This is the category that triggers most callbacks.
• BI-RADS 1 (Negative): No abnormalities. Routine screening interval continues.
• BI-RADS 2 (Benign): Definite benign findings such as cysts, intramammary lymph nodes, or calcified fibroadenomas. Routine screening continues.
• BI-RADS 3 (Probably Benign): Findings with >98% likelihood of benignity. Short-interval follow-up at 6 months is recommended.
• BI-RADS 4 (Suspicious): Divided into 4A (low suspicion, ~2–10% malignancy), 4B (intermediate suspicion, ~10–50%), and 4C (moderate concern, ~50–95%). Tissue biopsy is indicated.
• BI-RADS 5 (Highly Suggestive of Malignancy): Classic malignant features with ≥95% probability of cancer. Biopsy is mandatory.
• BI-RADS 6 (Known Biopsy-Proven Malignancy): Imaging performed for staging or treatment monitoring.

Key mammographic signs of malignancy

Radiologists evaluate five primary feature categories when assessing for malignancy:

• Masses: Malignant masses typically have irregular, spiculated, or ill-defined margins. Benign masses such as fibroadenomas are usually oval with circumscribed, smooth borders.
• Calcifications: Malignant calcifications are typically pleomorphic (varying shapes and sizes) or linear/branching (suggesting ductal involvement), often grouped in clusters or segmental distributions. Benign calcifications include popcorn (degenerating fibroadenomas), rim (fat necrosis), vascular, and coarse (involuting fibroadenomas).
• Architectural distortion: Focal disruption of normal breast architecture without a visible mass. This is a high-suspicion finding requiring biopsy even when subtle.
• Asymmetry: Global asymmetry (large area of increased density occupying at least one quadrant) is usually benign hormonal variation. Focal asymmetry (smaller, more defined) or developing asymmetry (new or enlarging) warrants targeted ultrasound and possible biopsy.
• Skin and nipple changes: Skin thickening, retraction, or nipple inversion may indicate underlying malignancy, particularly inflammatory breast cancer.

Common pitfalls and artefacts

Even with modern digital systems, mammography is susceptible to technical artefacts, patient-related factors, and interpretive challenges that can lead to false-positive callbacks or, more critically, false-negative misses. Understanding these pitfalls improves both patient communication and departmental quality assurance.

Technical artefacts

Common technical issues include:

• Deodorant and powder artefacts: Aluminium-based antiperspirants and talcum powders create dense white specks that mimic suspicious microcalcifications. These are usually recognisable by their location over the skin line or axilla, but they can obscure true underlying findings and necessitate repeat imaging after skin cleansing.
• Motion blur: Patient movement during the 1–2 second exposure degrades image sharpness. This is more common in 3D tomosynthesis when patients anticipate the arc motion. Firm, reassuring communication from the technologist and proper breathing instruction reduce motion rates.
• Skin fold artefacts: Improper positioning can trap a fold of skin, creating a linear density that may be mistaken for a mass or architectural distortion. Tangential views or repeat positioning with skin pull confirm the artefactual nature.
• Summation artefacts (2D only): Overlapping normal fibroglandular tissue can create pseudomasses on 2D views. Tomosynthesis effectively eliminates most summation artefacts by resolving tissue into thin slices.
• Implant-related challenges: Breast implants obscure approximately 25–35% of breast tissue on standard views. Implant-displaced Eklund views push the implant posteriorly to visualise anterior tissue, but these views are technically demanding and increase patient discomfort.

Interpretation traps

Even experienced radiologists encounter interpretive challenges:

• One-view-only findings: A density visible on only one projection may represent a summation artefact. True masses are visible on at least two orthogonal views. Tomosynthesis or spot compression views resolve this dilemma.
• Developing asymmetries: An asymmetry that is new or increasing compared with prior mammograms carries higher suspicion than a stable asymmetry documented over multiple years. Robust prior comparison is essential.
• Intramammary lymph nodes: These benign oval masses with a fatty hilum can mimic suspicious masses if the hilum is not clearly visible. Magnification views or ultrasound confirm the diagnosis.
• Radial scars / complex sclerosing lesions: These benign entities produce spiculated masses that are radiographically indistinguishable from invasive carcinoma. Core needle biopsy is required for definitive diagnosis.

Management implications

Mammographic findings directly guide clinical management pathways, from routine screening intervals to urgent biopsy referrals. A structured approach to callbacks, supplemental imaging, and follow-up surveillance ensures that patients receive appropriate care without unnecessary anxiety or delay.

The callback workflow

A callback (also called a diagnostic recall) occurs when the screening radiologist identifies a finding that requires additional imaging before a definitive BI-RADS category can be assigned. Callback rates vary by facility and patient population, ranging from 8–15% for 2D screening and 5–10% for combined 2D + 3D tomosynthesis^[8]. Critically, over 80% of callbacks resolve as benign after diagnostic workup, most commonly revealing cysts, overlapping tissue (summation artefacts), intramammary lymph nodes, or benign calcifications.

The diagnostic workup typically includes:

1. Spot compression views: Localised compression of a suspicious area separates tissue planes and clarifies whether a density represents a true mass or summation artefact.
2. Magnification views: High-resolution imaging of calcification clusters to assess morphology (pleomorphic vs benign) and distribution (segmental vs random).
3. Targeted breast ultrasound: Used to characterise masses as cystic (fluid-filled, benign) or solid (requiring biopsy), and to evaluate focal asymmetries.
4. 3D tomosynthesis cine review: Scrolling through reconstructed slices to localise findings in three dimensions and assess margins.

Supplemental screening for dense breasts

For women with Category C or D breast density, supplemental screening modalities improve cancer detection beyond mammography alone:

• Handheld screening breast ultrasound: Increases cancer detection by 2.8–4.4 per 1,000 women screened, though with a higher false-positive rate and increased callbacks^[9].
• Automated whole-breast ultrasound (ABUS): Standardised, operator-independent acquisition with computer-aided detection. Particularly useful in women with dense breasts who do not qualify for MRI.
• Contrast-enhanced mammography (CEM): Dual-energy technique using iodinated contrast to highlight tumour vascularity. Emerging as a cost-effective alternative to MRI with comparable sensitivity^[10].
• Breast MRI: The most sensitive supplemental modality, recommended for high-risk women (lifetime risk >20%) and those with known BRCA mutations. Detects an additional 15–20 cancers per 1,000 women screened^[11].

Biopsy pathways and follow-up surveillance

When imaging reveals a BI-RADS 4 or 5 lesion, percutaneous core needle biopsy is the standard of care. Stereotactic biopsy (using mammographic guidance) is preferred for calcifications and architectural distortions, while ultrasound-guided biopsy is used for sonographically visible masses. MRI-guided biopsy is reserved for lesions visible only on MRI.

Follow-up surveillance after benign biopsy or BI-RADS 3 assessment follows strict intervals:

• BI-RADS 3 (Probably Benign): 6-month mammogram, then 12-month, then return to routine screening if stable.
• Post-biopsy benign: 6–12 month mammogram to document biopsy site healing and exclude residual lesion.
• Post-lumpectomy: First post-treatment mammogram at 6 months, then annually.
• Post-mastectomy: Annual mammogram of the contralateral breast; MRI for high-risk surveillance if indicated.

Conclusion

Mammography remains the most validated, cost-effective, and widely accessible tool for early breast cancer detection. The transition from 2D digital mammography to 3D tomosynthesis represents a genuine paradigm shift, improving invasive cancer detection rates while simultaneously reducing the anxiety and cost burden of false-positive callbacks. For the approximately 50% of women with dense breast tissue, understanding the masking effect and the availability of supplemental screening modalities—ultrasound, contrast-enhanced mammography, and MRI—empowers informed shared decision-making with their healthcare providers.

Patient preparation is simple but consequential: avoiding skin products, scheduling at the optimal point in the menstrual cycle, and bringing prior imaging for comparison all contribute to a seamless, high-quality study. When callbacks occur, patients should be reassured that the vast majority resolve as benign, and that the callback system exists precisely to ensure that no suspicious finding is overlooked. As breast imaging technology continues to evolve with artificial intelligence-assisted detection and personalised risk-based screening protocols, the fundamental goal remains unchanged: finding breast cancer early, when treatment is most effective and survival is highest.