Comparison Of Digital Breast Tomosynthesis On Full-field Mammography On Assessment Of Breast Lesions

Objective:The comparison between digital breast tomosynthesis and full-field digital mammography focused on the two imaging methods difference in the sensitivity and the specificity of the diagnosis of breast lesions,and the difference in lesion BI-RADS classification,and the difference of diagnostic capabilities between the different dense of mammary glands of lesions. We expect to provide better clinical examination means and further improve the diagnostic level of breast tumor.Methods:All participant patients provided informed consent. From June 2014 to June 2015,we enrolled 251 breasts of 238 diagnostic patients with at least one breast mass finding found by clinical palpation. All cases were confirmed by biopsy or surgical pathology. All patients underwent examination consisted of the MLO and CC views, using combo automatic mode exposure, under the same conditions of compression,one exposure could be obtained FFDM and DBT images simultaneously. The FFDM and DBT images were retrospectively analyzed by two senior radiologists with more than 5 years of experience in breast imaging according to Breast Imaging Reporting and Data System(BI-RADS), recorded the reading results respectively. Chi-square and receiver operating characteristic(ROC) analysis were performed with the use of statistical software(SPSS, versionl9.0). Test level of a = 0.05, with P <0.05 to indicate a significant difference. Results:1.DBT and FFDM for the margin of mass: In DBT benign lesion grope,the margin of the mass was characterized clear, fuzzy and shadow were 83.5%,9.7%, 6.8%; In FFDM benign lesion grope,the margin of the mass was characterized clear, fuzzy and shadow were 43.7%,16.5%,38.3 %.There was statistically difference between the two methods(P=0.000). In DBT malignant lesion grope,the margin of the mass was characterized clear, fuzzy and shadow were 46.2%,43.7%,11.3%; In FFDM malignant lesion grope,the margin of the mass was characterized clear, fuzzy and shadow were 12.3%,64.2%,23.6 %.There was statistically difference between the two methods(P=0.000).2.In 125 cases of malignant lesion,38 cases were observed structural distortion in DBT,62 cases margin of mass display speculated,48 cases of mass were displayed increased blood flow,27 cases of mass could be observed the vessels penetration; In FFDM malignant lesions,there were 25 cases were observed structural distortion,48 cases margin of mass display speculated in DBT,39 cases of mass were displayed increased blood flow,9 cases of mass could be observed the vessels penetration. There was statistically difference between the two methods(P<0.05).3.DBT and FFDM diagnostic performance in different breast density:On mostly dense breast, DBT diagnostic sensitivity,specificity, PPV and NPV were 95.2%,73.5%,76.3% and 94.5%; FFDM diagnostic sensitivity,specificity,PPV and NPV were 93.3 %,55.6 %,65.3 % and 90.3 %. There was no significant difference in sensitivity(P>0.05), the difference of specificity and accuracy were statistically significant(P<0.05). The ROC curves showed the diagnostic accuracy of DBT was better than that of FFDM, the difference of AUC was 0.036.DBT diagnostic sensitivity, specificity, PPV and NPV were 100.0%,64.3%,80.0% and 100.0% on less dense breast,and FFDM diagnostic sensitivity,specificity,PPV and NPV diagnosis were 95.0%,57.8%,70.3%and 85.7%.The differences of sensitivity,specificity, and accuracy between two methods were not statistically significant. Area under the ROC showed the diagnostic accuracy of DBT was a litter better than that of FFDM, the difference of AUC was 0.0893.4.DBT and FFDM diagnostic performance in different size lesions: In 10 to 30 mm lesionsgroup,DBT diagnostic sensitivity,specificity, PPV and NPV were 97.4%,75.0%, 78.1% and 96.9 %;FFDM diagnostic sensitivity, specificity, PPV and NPV were 95.0 %,56.0%,66.1% and 90.3%. The differences of sensitivity between two methods is not statistically significant(P>0.05). The differences of specificity and accuracy between two methods were statistically significant(P<0.05). Area under the ROC showed the diagnostic accuracy of DBT was a litter better than that of FFDM, the difference of AUC was 0.054. In less than 10 mm lesions group,DBT diagnostic sensitivity, specificity,PPV and NPV were 85.7 %,70.1%,51.7% and 91.7%;FFDM diagnostic sensitivity,specificity,PPV and NPV were 85.7 %,58.1%,48.0 % and 90.0 %. The differences of sensitivity and specificity between two methods were not statistically significant(P>0.05). The differences of accuracy between two methods is statistically significant(P<0.05). In more than 30 mm lesions group,DBT diagnostic sensitivity, specificity, PPV and NPV were 97.1 %,62.5%, 84.6 % and 90.9 %; FFDM diagnostic sensitivity, specificity, PPV and NPV were 97.1 %, 37.5%, 76.7 % and 78.0%. The differences of sensitivity, specificity and accuracy between two methods were not statistically significant(P>0.05).5.DBT and FFDM overall diagnostic performance: The ROC showed that the diagnosis ability of breast lesions in DBT was higher than that in FFDM, the difference between the area under the curve was 0.0419. DBT diagnostic sensitivity,specificity, positive predictive value and negative predictive value were 96.0%,72.5%,76.9%and 95.5%; FFDM diagnostic sensitivity, specificity, positive predictive value and negative predictive value were 93.6%,54.2%,66.1 %and 89.8%.The difference of sensitivity between the two methods was not statistically significant(P =0.393), the difference of specificity was statistically significant(P =0.000), the difference of accuracy was statistically significantly(P=0.003).6. DBT and FFDM BI-RADS of benign and malignant breast lesions classification comparison: in benign lesion, DBT and FFDM diagnosis were mainly distributed in the BI-RADS 2, BI-RADS 3 and BI-RADS 4 a class, DBT and FFDM in breast benign lesions BI- RADS classification differences statistically significant(χ2 = 14.616, P = 14.616). In malignant lesions group, diagnosis are mainly distributed in the BI-RADS 4c and BI-RADS 5 classes,DBT and FFDM in breast malignant lesions there was no statistically significant difference in the BI-RADS classification(χ2 = 8.978, P = 8.978). Conclusion:1.DBT can reduce the overlap between organizations that the margin of lesion can be showed more clearly, especially for speculated, structural distortions and vascular penetration. DBT can improve the diagnostic accuracy of benign and malignant breast masses. Further, because of the advantages of the margins of masses display, DBT in measurement of masses size is also superior FFDM. DBT has more advantages of the display of the margins、structural distortions and also vascular penetration.2.In dense glands group,DBT is less affected by the glands overlapping then FFDM, the masses covered by glands are shown.DBT is able to show that the structure of the lesion, edge and blood flow to show more clearly,improves specificity and accuracy.In non dense glands group, DBT have no advantage of the diagnosis.3.In 10-30 mm group, DBT have more advantages of the diagnosis than FFDM.4.DBT diagnostic capacity is higher than FFDM,sensitivity,specificity and accuracy were higher than FFDM,in particular,specificity and accuracy are improved obviously.5.The classification of DBT compared with FFDM,DBT is more accurate,more benign lesions are classified as BI-RADS category 2 and BI-RADS category 3; more malignant lesions are classified as BI-RADS category 4 and BI-RADS category 5.