Scientific Session 03 — Breast - Digital Breast TomosythesisMonday, May 1, 2017
1108. Analysis of Micro-Calcifications Detected on Screening Synthetic 2D Mammography: Are They All Real?
Sharma K, Jalab F*, Harkins K. St. Vincent's Medical Center, Bridgeport, CT
Address correspondence to K. Sharma (firstname.lastname@example.org)
Objective: We aim to assess the accuracy and possible impact on patient recall rates using synthetic 2D mammography in combination with digital breast tomosynthesis (DBT). Specifically, we have noticed an artifact (pseudocalcifications) on synthetic 2D mammograms, ultimately increasing the number of patients recalled for further evaluation of possible calcifications noted on screening mammograms. Benefits of DBT in screening and diagnostic mammography have been well established . However, the estimated radiation dose per view for combined 2D full field digital mammography (FFDM) and two-view DBT is 3.53 mGy, and it is 1.58 mGy per view for FFDM alone. The patient dose can be reduced to 1.95 mGy by combining synthetic 2D mammography and DBT. Despite the recent decision of the Food and Drug Administration allowing synthetic 2D mammograms to replace the standard combined 2D views performed in addition to the DBT, few practices use synthetic 2D mammograms. The supporting quality of synthetic 2D mammography is not well established, likely resulting in its underutilization.
Materials and Methods: This is a retrospective study that included 109 consecutive women aged 34–80 years who were recalled between September 2015 to July 2016 for further evaluation of microcalcifications detected on screening synthetic 2D mammography and DBT. All 109 patients subsequently underwent diagnostic FFDM including magnification views on follow-up. The magnification views were considered to be conclusive. Follow-up studies that demonstrated no calcifications on follow-up magnification views were categorized as BI-RADS category 1 on FFDM. This group of patients was considered to have a false-positive finding (pseudocalcifications) on the synthetic 2D mammogram. The final BI-RADS assessment on FFDM was correlated with breast density (assigned according to standard mammography lexicon) and patient age using ANOVA.
Results: Of the 109 patients found to have suspicious calcification on screening synthetic 2D mammography, 14 patients did not demonstrate any calcification on subsequent diagnostic FFDM, giving a false-positive rate of 12.8%. Further, among different breast density categories, the false-positive rate was 14.3% among fatty, heterogeneous, and extremely dense categories and 9.4% in the scattered breast density category. There was no statistically significant difference among patients with different breast density categories (p = 0.605) or different age groups (p = 0.087). Thirty-one of 109 patients (28.4%) were found to have benign calcifications (BI-RADS category 2). Sixty-four of 109 patients (58.7%) were assigned BI-RADS category 3, 4, or 5 and needed follow-up, biopsy, or surgical consult.
Conclusion: Synthetic 2D mammography offers the benefit of radiation dose reduction, but we found a high number of false-positive microcalcifications with use of synthetic 2D mammography when compared with FFDM. These calcifications seen on synthetic 2D mammography can be termed pseudocalcifications. A mechanism for pseudocalcifications is unclear; however, they may be related to reconstruction algorithms used for synthetic 2D mammography. With increasing availability and use of synthetic 2D mammography, it is essential for the radiologist to be aware of this potential pitfall associated with this technique. Further large-scale studies are needed to better characterize and understand the mechanics of pseudocalcifications.