Diagnostic value of digital breast tomosynthesis and synthetic images in patients with breast cancer.

Abstract

Different imaging techniques are used in the diagnosis of breast cancer. The low sensitivity of mammography to detect cancer in the dense breast parenchyma and the lack of standard application of digital breast tomosynthesis (DBT) are some of the problems. Therefore, breast cancer imaging techniques should be compared in terms of conspicuity and characterization of lesions. Full-field digital mammography (DM) and synthetic mammography (SM) which are obtained from the slices of digital breast tomosynthesis (DBT) give similar results in terms of conspicuity and characterization of the lesions in detecting breast cancer. In this retrospective study, 47 women diagnosed with breast cancer were included in the study. DM, SM, and DBT images were evaluated by scoring the conspicuity of the index lesion in the parenchyma and its characterization in terms of contour and shape with a 4-point scale. In addition, the conspicuity of the lesions in relation to lesion size and breast density was examined with these three techniques. There is no significant difference between DM and SM techniques for index lesion conspicuity and characterization; however, the imaging score of DBT is significantly higher than other techniques for the conspicuity and characterization of the lesions. In terms of the conspicuity of the lesions in relation to lesion size, DM and SM techniques show significant difference according to the size of the lesion, whereas the DBT technique did not show significant difference. While mammography type is a determinant of lesion conspicuity in only DM and SM techniques, conspicuity findings do not differ significantly in the DBT technique. In conclusion, it was shown that standard images and SM images obtained from DBT did not differ significantly in terms of conspicuity and characterization of lesions. Thus, DBT is significantly superior to the DM and SM images. While the DM and SM images are more successful in showing large lesions and lesion detection in nondense breasts, DBT images were not affected by lesion size and breast density.