[Syllabus]: 2-D Mammography Synthesized from Tomosynthesis: Strengths, Pitfalls, Artifacts

Sona A. Chikarmane, MD

Sona A. Chikarmane, MD(bio)

  • Associate Professor of Radiology
  • Boston University Chobanian & Avedisian School of Medicine

Introduction

This presentation reviews synthetic mammography (SM)—a two-dimensional image reconstructed from digital breast tomosynthesis (DBT)—with emphasis on image formation, radiation dose, clinical performance, artifacts, strengths, and practical pitfalls. As SM is replacing full-field digital mammography (FFDM) in many practices to reduce dose while preserving diagnostic performance, radiologists must understand its physics, characteristic artifacts, interpretive strategies, and workflow implications to maintain high-quality breast imaging and patient safety.

Synthetic Mammography in the Screening Workflow

Synthetic mammography generates a 2D projection image from a DBT acquisition, replacing the separately acquired FFDM in a combined exam. The traditional “combo” workflow acquires DBT plus FFDM; the SM workflow acquires DBT alone and synthesizes the 2D view for interpretation alongside the DBT slices.

Key Points

  • SM is a 2D image reconstructed from DBT, not an independently acquired exposure.
  • Reading paradigm: interpret DBT plus SM together; SM should not be used without its DBT counterpart.
  • Many centers are adopting DBT+SM to eliminate the extra FFDM exposure and shorten exams.

Image Formation Physics and Algorithmic Design

A standard DBT acquisition is reconstructed into approximately 1-mm slices. Vendor-specific algorithms identify calcification-like and lesion-like features across slices and generate a synthesized 2D image, giving extra weighting to suspicious regions relative to normal fibroglandular and adipose tissue to enhance conspicuity.

Key Points

  • Reconstruction emphasizes high-frequency bright foci (calcifications) and lesion texture to improve visibility.
  • Weighting strategies are vendor-specific and can accentuate noise and structural overlap.
  • The design goal is improved detectability of masses and calcifications compared with a simple average projection.

Radiation Dose Framework and Rationale

Under MQSA, the per-view dose limit is 3 mGy using a standard 4.2-cm, 50% glandularity phantom. DBT+FFDM increases dose approximately 2.25–2.5× compared with FFDM alone. Replacing FFDM with SM removes the second exposure, yielding a DBT-equivalent dose for the combined exam.

Key Points

  • SM adds no additional radiation—dose equals DBT alone.
  • Reported average glandular dose reductions replacing FFDM with SM: ~39–45% in combined exams.
  • Dose reduction is particularly meaningful for annual screening and younger patients.

Clinical Performance Evidence

Multiple studies demonstrate that DBT+SM performs comparably to DBT+FFDM and to FFDM alone for cancer detection and recall metrics. Multicenter and retrospective data show no significant differences in sensitivity, specificity, or calcification visibility between SM and FFDM when used appropriately with DBT.

Key Points

  • Gilbert et al: DBT+SM sensitivity and specificity comparable to FFDM and DBT+FFDM.
  • Choi et al (T1 cancers): no significant differences in cancer detection, visibility, or calcifications between SM and FFDM.
  • Zuley et al: SM alone or SM+DBT comparable to FFDM alone or FFDM+DBT.
  • SM should always be interpreted with DBT for optimal performance.

Artifact: Subcutaneous Tissue Blurring (“Bright Band”) and Burned Skin

SM can show a superficial bright band with reduced skin/subcutaneous detail, attributed to limited information at the extremes of the curved compressed breast during tomosynthesis reconstruction. In dense/thick breasts, high exposure may saturate the detector at the skin line (“burned skin”), further degrading skin definition.

Key Points

  • Expect decreased visualization of superficial lesions (e.g., subcutaneous calcifications) on SM.
  • Mechanisms: fewer informative rays at the compression plate and bucky; detector saturation in dense tissue.
  • Mitigation: rely on DBT slices to interrogate the skin/subcutaneous layer carefully.

Artifact: Decreased Axillary Resolution and High-Density Shoulder Effects

SM may demonstrate hazy axillary detail, especially on MLO/lateral views, due to dominance of homogeneous pectoralis soft tissue in the synthesis. High-density structures (e.g., shoulder) within the tomosynthesis arc can amplify regional artifacts.

Key Points

  • Axillary nodes and pathology can be obscured on SM; DBT slices are crucial for axillary assessment.
  • Homogeneous pectoralis density reduces local contrast; structural artifacts may be more conspicuous on SM.
  • Carefully scrutinize the axilla on DBT, not SM alone.

Artifact: Pseudo-calcifications

The accentuation algorithm enhances bright speckle and structural overlap (e.g., ligaments), creating calcification-like foci that are not real. These may be seen on SM but absent on DBT and FFDM.

Key Points

  • Troubleshooting: scroll DBT slices; if absent or only a single speck, consider pseudo-calcifications.
  • Confirm on the orthogonal view; true calcifications typically appear on both views and DBT.
  • Anticipate a learning curve; early use may transiently increase callbacks.

Artifact: Metallic Foreign Bodies and Skin Markers

Metallic clips and markers produce streak artifacts on SM analogous to CT beam-hardening, degrading local contrast and obscuring the tissue–metal interface.

Key Points

  • Expect loss of detail adjacent to clips; underlying masses may appear less conspicuous on SM than FFDM.
  • Use DBT slices to evaluate tissue around metal; consider artifact-reducing markers that minimize high-attenuation effects.
  • Vendor techniques to reduce metal artifacts are evolving.

Motion Detection and Quality Assurance with DBT/SM

Motion during DBT (e.g., patient sneezing) may be subtle or absent on SM but visible on DBT projection cine or slices. Because SM is synthesized from motion-degraded DBT data, detection relies on DBT review rather than SM.

Key Points

  • Review DBT projection images in cine mode for “bouncing” at the inframammary fold, axillary skin (MLO), and cleavage (CC).
  • Clues include non-vertical “slinky” artifact orientation and unexpected blurring of masses/microcalcifications on slices.
  • FFDM can mask DBT motion absence in an SM workflow; rigorous DBT QA review is essential.

Operational Advantages: Acquisition Time, Throughput, and Patient Experience

Eliminating the separate FFDM acquisition shortens exams and can improve workflow. Reported average exam-time savings approximate one minute per examination when omitting FFDM in a system with ~4 seconds per view acquisition.

Key Points

  • Reduced acquisition time and radiographer workload per patient increase throughput capacity.
  • Shorter compression time may improve patient comfort and reduce motion-related technical repeats.
  • DBT+SM has the same acquisition time as DBT alone, versus longer time for DBT+FFDM.

Diagnostic Strength: Calcifications

SM often renders calcifications sharper and more conspicuous than FFDM due to algorithmic weighting of bright, high-frequency features. Multiple case examples demonstrated improved depiction of grouped pleomorphic and amorphous calcifications corresponding to DCIS.

Key Points

  • Calcification clarity and edge definition are frequently enhanced on SM.
  • Enhanced conspicuity supports detection of DCIS; correlate with DBT to confirm morphology and distribution.
  • Balance the benefit against the risk of pseudo-calcifications; verify on DBT and orthogonal view.

Diagnostic Strength: Spiculated Margins

SM can accentuate spicules and margin detail of malignant masses relative to FFDM, improving perception of irregularity and radiating lines.

Key Points

  • Spiculated margins may appear longer and sharper on SM, aiding recognition of invasive carcinomas.
  • Use SM to localize suspicious margins and DBT to analyze the spiculations in depth.
  • Correlate with ultrasound for targeted evaluation and biopsy planning.

Diagnostic Strength: Architectural Distortion

Architectural distortion may be better appreciated on SM, showing alternating radiating lines (spoke-wheel pattern) that are subtle or occult on FFDM alone.

Key Points

  • SM can reveal distortion patterns not evident on FFDM; DBT confirms the distortion plane and extent.
  • Architectural distortion on SM warrants careful DBT correlation and targeted sonography.
  • This enhanced visualization supports early cancer detection.

Pitfalls, Callbacks, and the Learning Curve

Initial adoption of SM can temporarily increase callbacks, particularly for pseudo-calcifications and superficial findings obscured by the bright-band artifact. Systematic use of DBT correlation and orthogonal confirmation reduces unnecessary recalls over time.

Key Points

  • Always confirm SM findings on DBT and the other view before recalling for calcifications.
  • Pay extra attention to the skin/subcutaneous layer on DBT where SM shows bright-band blurring.
  • Establish internal consensus criteria and education to standardize pseudo-calcification recognition.

Coding and CAD Considerations

There is no separate CPT code or standard additional reimbursement specific to SM; SM and FFDM are both planar images for coding purposes. CAD solutions for SM exist, but comparative performance data versus FFDM CAD remain limited.

Key Points

  • Code SM as a planar mammographic image; reimbursement typically aligns with institutional DBT policies.
  • CAD is available for SM, though evidence comparing SM-CAD to FFDM-CAD is sparse.
  • Stay current with payer and regulatory updates regarding SM reimbursement.

Practical Reading Tips and Protocol Recommendations

Adopting consistent reading protocols enhances accuracy and efficiency using SM with DBT.

Key Points

  • Always interpret SM with DBT; never rely on SM alone for diagnostic decisions.
  • For suspected motion, review DBT projections in cine and scrutinize skin lines in standard regions.
  • Use artifact-aware strategies: check axilla on DBT, investigate metal-adjacent tissues on DBT, and verify calcifications across views.
  • Consider artifact-reducing skin/scar markers and keep abreast of vendor algorithm updates that may impact artifact profiles.

Conclusion

Synthetic mammography provides a dose-neutral 2D companion to DBT that can replace FFDM in many screening workflows without compromising cancer detection or recall metrics. Its algorithmic enhancement of calcifications, spiculations, and architectural distortion confers diagnostic strengths, while characteristic artifacts—subcutaneous bright band, axillary blurring, pseudo-calcifications, metal streaks, and motion invisibility—require informed interpretation strategies anchored in DBT review. With standardized protocols, attention to artifact recognition, and an appreciation of operational benefits, DBT+SM can optimize breast imaging quality, safety, and efficiency.