|
KMID : 1114620200170020093
|
|
Journal of the Korean Society for Breast Screening
2020 Volume.17 No. 2 p.93 ~ p.99
|
|
|
Agreement in Breast Density Assessment between Two-Dimensional Synthetic Mammography and Digital Breast Tomosynthesis
|
|
Choi Ji-Ae
Eun Na-Lae
Youk Ji-Hyun
Son Eun-Ju
Gweon Hye-Mi
Kim Jeong-Ah
|
|
|
Abstract
|
|
|
Purpose: To evaluate the agreement of mammographic breast density between digital breast tomosynthesis (DBT) and two-dimensional synthetic mammography (2D-SM).
Materials and Methods: For mammographic breast density, two radiologists specialized in breast imaging independently reviewed 231 DBT and corresponding 2D-SM images obtained from 231 women in July 2020 according to breast density category in BI-RADS 5th edition. A weighted kappa value was calculated to assess the degree of agreement in density category between observers or between modalities, and the chi-square test was performed to compare the agreement rate by dense breast.
Results: Interobserver agreement was fair for breast density assessed in both DBT (κ = 0.551) and 2DSM (κ = 0.543). Agreement for breast density between two modalities was moderate in observer 1 (κ = 0.741) and good in observer 2 (κ = 0.812). For observer 1, DBT showed significantly lower rate of category C and higher rate of category B and D than DBT (P = 0.046) but for observer 2, the distribution of density category was not significantly different between modalities (P = 0.550).
For the dichotomized category by dense breast, both modalities showed no significant difference in agreement rate between observers (DBT, 93.1%; 2D-SM, 92.2%; P = 0.8586) and both observers showed no significant difference in agreement rate between modalities (observer 1, 93.9%; observer 2, 97.4%; P = 0.1095).
Conclusion: According to BI-RADS breast density, 2D-SM and DBT showed good agreement for dense breast.
|
|
|
KEYWORD
|
|
|
Breast density, Digital breast tomosynthesis, Synthetic 2D mammography
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|