Automated Classification of Benign and Malignant Proliferative Breast Lesions.
Automated Classification of Benign and Malignant Proliferative Breast Lesions.
复制标题
良性和恶性增殖性乳腺病变的自动分类。
DOI：
10.1038/s41598-017-10324-y
复制
发表时间：
2017
影响因子：
4.6
通讯作者：
Beck,AndrewH
中科院分区：
文献类型：
--
作者：
Radiya-Dixit,Evani;Zhu,David;Beck,AndrewH
乳腺病变的错误分类可能导致癌症进展或不必要的化疗。自动分类工具被视为有希望减少此类错误的第二意见提供者。我们开发了预测算法，可自动将乳腺病变分类为良性普通导管增生 (UDH) 或恶性导管原位癌 (DCIS)。从两家医院诊断的乳腺活检图像中，我们使用 Donget al. (2014) 的计算工具进行细胞核识别和特征提取，获得了 392 个生物标志物。我们实现了六种机器学习模型，并通过减少预测方差、提取主动特征和组合多种算法来增强它们。我们使用受试者工作特征（ROC）曲线的曲线下面积（AUC）进行性能评估。我们表现最好的模型是一种由两种逻辑回归算法组成的主动特征提取 (CAFE) 组合模型，在对一家医院的数据进行训练并在另一家医院的样本上进行测试时，获得了 0.918 的 AUC，比 Donget 等人的 0.858 的 AUC 有统计学上的显着改进。病理学家可以通过将其用作无偏见的验证器来显着改善他们的诊断。将来，我们的工作还可以作为区分低级别和高级别 DCIS 的有价值的方法。
Misclassification of breast lesions can result in either cancer progression or unnecessary chemotherapy. Automated classification tools are seen as promising second opinion providers in reducing such errors. We have developed predictive algorithms that automate the categorization of breast lesions as either benign usual ductal hyperplasia (UDH) or malignant ductal carcinomain situ(DCIS). From diagnosed breast biopsy images from two hospitals, we obtained 392 biomarkers using Donget al.’s (2014) computational tools for nuclei identification and feature extraction. We implemented six machine learning models and enhanced them by reducing prediction variance, extracting active features, and combining multiple algorithms. We used the area under the curve (AUC) of the receiver operating characteristic (ROC) curve for performance evaluation. Our top-performing model, a Combined model with Active Feature Extraction (CAFE) consisting of two logistic regression algorithms, obtained an AUC of 0.918 when trained on data from one hospital and tested on samples of the other, a statistically significant improvement over Donget al.’s AUC of 0.858. Pathologists can substantially improve their diagnoses by using it as an unbiased validator. In the future, our work can also serve as a valuable methodology for differentiating between low-grade and high-grade DCIS.