Medical Radiation Physics, Department of Translational Medicine, Faculty of Medicine, Lund University, Skåne University Hospital, 205 02, Malmö, Sweden.
Department of Diagnostic Imaging and Nuclear Medicine, Graduate School of Medicine, Kyoto University, 54 Shogoin-Kawaharacho Sakyo-ku, Kyoto, 606-8507, Japan.
Eur Radiol. 2018 Aug;28(8):3194-3203. doi: 10.1007/s00330-017-5297-7. Epub 2018 Feb 19.
To compare breast density measured on digital breast tomosynthesis (DBT) (BI-RADS-based breast composition and fully-automatic estimation) and magnetic resonance imaging (MRI) (BI-RADS amount of fibroglandular tissue), and to evaluate the diagnostic performance in terms of sensitivity and specificity of DBT and MRI in a predominantly dense breast population.
Between 2015 and 2016, 152 women with 103 breast malignancies, who underwent 3-T breast MRI and DBT within 2 months' time, were enrolled in this study. Breast composition/fibroglandular tissue and findings on DBT (two readers) and MRI were reported using BI-RADS 5th edition. Digital mammography images were analysed for breast percent density (PD) using the Libra software tool.
A majority of women had dense breasts as categorised by breast composition c (heterogeneously dense) (68%) and d (extremely dense) (15%). The mean PD was 44% (range, 18-89%) and the correlation between breast composition and PD was r = 0.6. The diagnostic performance of MRI was significantly higher compared to DBT for one reader as described by the area under the receiver operating characteristic (ROC) curve (p = 0.004) and of borderline significance for the other reader (p = 0.052).
MRI had higher diagnostic performance than DBT in a dense breast population in the tertiary setting.
• MRI had higher diagnostic performance than DBT in a dense breast population • Diagnostic performance of DBT was comparable to MRI in women with fatty breasts • MRI was superior to DBT in preoperative breast cancer size assessment.
比较数字乳腺断层摄影术(DBT)(基于 BI-RADS 的乳腺成分和全自动评估)和磁共振成像(MRI)(BI-RADS 纤维腺体组织量)测量的乳腺密度,并评估在以致密乳腺为主的人群中,DBT 和 MRI 的诊断性能,包括敏感性和特异性。
2015 年至 2016 年间,共纳入 152 例 103 例乳腺恶性肿瘤患者,这些患者在 2 个月内接受了 3T 乳腺 MRI 和 DBT 检查。使用 BI-RADS 第 5 版报告乳腺成分/纤维腺体组织和 DBT(两位读者)及 MRI 的发现。使用 Libra 软件工具对数字乳腺 X 线照片进行乳腺百分比密度(PD)分析。
大多数女性的乳腺成分被归类为 c(不均匀致密)(68%)和 d(极度致密)(15%),表现为致密乳腺。平均 PD 为 44%(范围 18-89%),乳腺成分与 PD 之间的相关性为 r = 0.6。一位读者的 ROC 曲线下面积(AUC)显示 MRI 的诊断性能明显高于 DBT(p = 0.004),另一位读者的 AUC 显示 MRI 的诊断性能具有边界意义(p = 0.052)。
在三级环境中,MRI 对致密乳腺人群的诊断性能优于 DBT。
MRI 在致密乳腺人群中的诊断性能优于 DBT。
在脂肪性乳腺的女性中,DBT 的诊断性能与 MRI 相当。
MRI 在术前乳腺癌大小评估方面优于 DBT。