Evaluation of microcalcification contrast in clinical images for digital mammography and synthetic mammography

Synthetic Mammography: Benefits, Drawbacks, and Pitfalls

Digital breast tomosynthesis (DBT) allows three-dimensional assessment of breast tissue; however, DBT requires a two-dimensional (2D) image for comparison with prior mammograms and accurate interpretation of calcifications. Traditionally, full-field digital mammography (FFDM) has been performed after the DBT image acquisition. Synthetic mammography (SM), the 2D reconstruction of the tomosynthesis slice dataset, has been designed to replace FFDM. Advantages of SM include decreased image acquisition time and decreased radiation exposure, with maintained or improved screening performance metrics. Because SM algorithms give extra weight to lesion-like characteristics (eg, calcifications and architectural distortions), they may enable increased visibility of these characteristics relative to that at FFDM. Although SM algorithms were designed to improve lesion identification, they have led to varied outcomes in studies reported in the literature. Compared with FFDM, SM has been reported to be associated with a higher false-positive rate for calcifications, decreased conspicuity of asymmetries, lower breast density assessments, and imaging artifacts (eg, metallic artifact, bright-band artifact, blurring of the axilla, and truncation artifact). The authors review the literature on SM, including its implementation, benefits, and artifacts. ©RSNA, 2023 Quiz questions for this article are available through the Online Learning Center.

Comparisons between artificial intelligence computer-aided detection synthesized mammograms and digital mammograms when used alone and in combination with tomosynthesis images in a virtual screening setting

PURPOSE

To compare the reader performance of artificial intelligence computer-aided detection synthesized mammograms (AI CAD SM) with that of digital mammograms (DM) when used alone or in combination with digital breast tomosynthesis (DBT) images.

MATERIALS AND METHODS

This retrospective multireader (n = 4) study compared the reader performances in 388 cases (84 cancer, 83 benign, and 221 normal or benign cases). The overall accuracy of the breast-based assessment was determined by four radiologists using two sequential reading modes: DM followed by DM + DBT; and AI CAD SM followed by AI CAD SM + DBT. Each breast was rated by each reader using five-category ratings, where 3 or higher was considered positive. The area under the receiver-operating characteristic curve (AUC) and reading time were evaluated.

RESULTS

The mean AUC values for DM, AI CAD SM, DM + DBT, and AI CAD SM + DBT were 0.863, 0.895, 0.886, and 0.902, respectively. The mean AUC of AI CAD SM was significantly higher (P < 0.0001) than that of DM. The mean AUC of AI CAD SM + DBT was higher than that of DM + DBT (P = 0.094). A significant reduction in the reading time was observed after using AI CAD SM + DBT when compared with that after using DM + DBT (P < 0.001).

CONCLUSION

AI CAD SM + DBT might prove more effective than DM + DBT in a screening setting because of its lower radiation dose, noninferiority, and shorter reading time compared to DM + DBT.

Artificial intelligence computer-aided detection enhances synthesized mammograms: comparison with original digital mammograms alone and in combination with tomosynthesis images in an experimental setting

BACKGROUND

It remains unclear whether original full-field digital mammograms (DMs) can be replaced with synthesized mammograms in both screening and diagnostic settings. To compare reader performance of artificial intelligence computer-aided detection synthesized mammograms (AI CAD SMs) with that of DM alone or in combination with digital breast tomosynthesis (DBT) images in an experimental setting.

METHODS

We compared the performance of multireader (n = 4) and reading multicase (n = 388), in 84 cancers, 83 biopsy-proven benign lesions, and 221 normal or benign cases with negative results after 1-year follow-up. Each reading was independently interpreted with four reading modes: DM, AI CAD SM, DM + DBT, and AI CAD SM + DBT. The accuracy of probability of malignancy (POM) and five-category ratings were evaluated using areas under the receiver operating characteristic curve (AUC) in the random-reader analysis.

RESULTS

The mean AUC values based on POM for DM, AI CAD SM, DM + DBT, and AI CAD SM + DBT were 0.871, 0.902, 0.895, and 0.909, respectively. The mean AUC of AI CAD SM was significantly higher (P = 0.002) than that of DM. For calcification lesions, the sensitivity of SM and DM did not differ significantly (P = 0.204). The mean AUC for AI CAD SM + DBT was higher than that of DM + DBT (P = 0.082). ROC curves based on the five-category ratings showed similar proximity of the overall performance levels.

CONCLUSIONS

AI CAD SM alone was superior to DM alone. Also, AI CAD SM + DBT was superior to DM + DBT but not statistically significant.

Synthetic Mammography: Review of Benefits and Drawbacks in Clinical Use

Digital breast tomosynthesis (DBT) has been widely adopted as a breast cancer screening tool, demonstrating decreased recall rates and other improved screening performance metrics when compared to digital mammography (DM) alone. Drawbacks of DBT when added to 2D DM include the increased radiation dose and longer examination time. Synthetic mammography (SM), a 2D reconstruction from the tomosynthesis slices, has been introduced to eliminate the need for a separate acquisition of 2D DM. Data show that the replacement of 2D DM by SM, when used with DBT, maintains the benefits of DBT, such as decreased recall rates, improved cancer detection rates, and similar positive predictive values. Key differences between SM and 2D DM include how the image is acquired, assessment of breast density, and visualization of mammographic findings, such as calcifications. Although SM is approved by the Food and Drug Administration and has been shown to be non-inferior when used with DBT, concerns surrounding SM include image quality and artifacts. The purpose of this review article is to review the benefits, drawbacks, and screening performance metrics of SM versus DBT.