Jong R, Fishell E, Little L, Lockwood G, Boyd N F
Department of Radiology, Mount Sinai Hospital, Toronto, Ontario, Canada.
Eur J Cancer Prev. 1996 Aug;5(4):281-6. doi: 10.1097/00008469-199608000-00008.
We have assessed the agreement of radiologists who quantitatively estimated the extent of various mammographic features of the breast. These features are of potential relevance to the estimation of the future risk of developing breast cancer. After preparation, in which various mammographic appearances and their classification were discussed and agreed upon, two radiologists, experienced in mammography, independently estimated the extent of various types of radiological density in 120 sets of mammograms, each comprised of two mediolateral and two craniocaudal views. The results showed a high level of agreement between the two radiologists in the classification of total mammographic densities (r = 0.89). For classification of specific types of density, agreement was best for homogeneous densities (r = 0.79), more intermediate for nodular densities (r = 0.71) and worst for linear densities (r = 0.48). These results show that radiologists can, with suitable preparation, identify total breast density, a mammographic sign relevant to a woman's risk of later developing breast cancer, with a high degree of agreement.
我们评估了对乳腺各种乳房X线特征范围进行定量估计的放射科医生之间的一致性。这些特征与未来患乳腺癌风险的估计可能相关。在进行准备工作时，讨论并商定了各种乳房X线表现及其分类，两位有乳房X线摄影经验的放射科医生独立估计了120组乳房X线照片中各种类型放射密度的范围，每组照片包括两张内外侧位和两张头尾位视图。结果显示，两位放射科医生在乳房X线总体密度分类方面具有高度一致性（r = 0.89）。对于特定类型密度的分类，均匀密度的一致性最佳（r = 0.79），结节状密度的一致性居中（r = 0.71），线状密度的一致性最差（r = 0.48）。这些结果表明，经过适当准备后，放射科医生能够高度一致地识别乳房总体密度，这是一种与女性日后患乳腺癌风险相关的乳房X线征象。
