Department of Radiology, University of Washington School of Medicine, Seattle Cancer Care Alliance, 1144 Eastlake Ave E, Room LG2-211, Seattle, WA 98109 (H.R., D.S.H., S.H.C., S.C.P., C.I.L.); Department of Electrical and Computer Engineering, University of California, Los Angeles, Los Angeles, Calif (A.A., M.v.d.S.); and Departments of Applied Mathematics and Theoretical Physics and Public Health, University of Cambridge, Cambridge, England (M.v.d.S.).
Radiol Imaging Cancer. 2020 Jul 10;2(4):e190099. doi: 10.1148/rycan.2020190099.
To identify patient and tumor features that predict true-positive, false-positive, and negative breast preoperative MRI outcomes.
Using a breast MRI database from a large regional cancer center, the authors retrospectively identified all women with unilateral breast cancer who underwent preoperative MRI from January 2005 to February 2015. A total of 1396 women with complete data were included. Patient features (ie, age, breast density) and index tumor features (ie, type, grade, hormone receptor, human epidermal growth factor receptor type 2/, Ki-67) were extracted and compared with preoperative MRI outcomes (ie, true positive, false positive, negative) using univariate (ie, Fisher exact) and multivariate machine learning approaches (ie, least absolute shrinkage and selection operator, AutoPrognosis). Overall prediction performance was summarized using the area under the receiver operating characteristic curve (AUC), calculated using internal validation techniques (bootstrap and cross-validation) to account for model training.
At the examination level, 181 additional cancers were identified among 1396 total preoperative MRI examinations (median patient age, 56 years; range, 25-94 years), resulting in a positive predictive value for biopsy of 43% (181 true-positive findings of 419 core-needle biopsies). In univariate analysis, no patient or tumor feature was associated with a true-positive outcome ( > .05), although greater mammographic density ( = .022) and younger age (< 50 years, = .025) were associated with false-positive examinations. Machine learning approaches provided weak performance for predicting true-positive, false-positive, and negative examinations (AUC range, 0.50-0.57).
Commonly used patient and tumor factors driving expert opinion for the use of preoperative MRI provide limited predictive value for determining preoperative MRI outcomes in women. © RSNA, 2020See also the commentary by Grimm in this issue.
确定预测乳腺术前 MRI 阳性、假阳性和阴性结果的患者和肿瘤特征。
利用一家大型地区癌症中心的乳腺 MRI 数据库，作者回顾性地识别了 2005 年 1 月至 2015 年 2 月期间接受单侧乳腺癌术前 MRI 的所有女性患者。共纳入 1396 例数据完整的女性患者。提取患者特征（即年龄、乳腺密度）和肿瘤指标特征（即类型、分级、激素受体、人表皮生长因子受体 2/，Ki-67），并与术前 MRI 结果（即真阳性、假阳性、阴性）进行比较，采用单变量（即 Fisher 确切检验）和多变量机器学习方法（即最小绝对收缩和选择算子、AutoPrognosis）。使用内部验证技术（自举和交叉验证）计算接收器工作特征曲线下面积（AUC），以考虑模型训练，从而总结整体预测性能。
在检查层面上，在 1396 例术前 MRI 检查中，又发现了 181 例额外的癌症（中位患者年龄为 56 岁；范围为 25-94 岁），导致活检的阳性预测值为 43%（419 例空心针活检中有 181 例真阳性发现）。在单变量分析中，没有患者或肿瘤特征与真阳性结果相关（>.05），尽管更大的乳腺密度（=.022）和更年轻的年龄（<50 岁，=.025）与假阳性检查相关。机器学习方法对预测真阳性、假阳性和阴性检查的性能较差（AUC 范围为 0.50-0.57）。
推动专家意见使用术前 MRI 的常用患者和肿瘤因素，对确定女性术前 MRI 结果的预测价值有限。 ©2020RSNA，见本期 Grimm 的评论。
