Deep learning of longitudinal mammogram examinations for breast cancer risk prediction.
Deep learning of longitudinal mammogram examinations for breast cancer risk prediction.
复制标题
纵向乳房X光检查的深度学习用于乳腺癌风险预测。
DOI：
10.1016/j.patcog.2022.108919
复制
发表时间：
2022
影响因子：
8
通讯作者：
Wu,Shandong
中科院分区：
文献类型：
--
作者：
Dadsetan,Saba;Arefan,Dooman;Berg,WendieA;Zuley,MargaritaL;Sumkin,JulesH;Wu,Shandong
数字乳房X线照片中的信息已被证明与患乳腺癌的风险有关。纵向乳腺癌筛查乳房X线检查可能携带时空信息，可以提高乳腺癌的风险预测。目前还没有设计深度学习模型来通过多次检查捕获这种时空信息以预测风险。在这项研究中，我们提出了一种新的深度学习结构LRP-NET，用于捕获乳腺组织在多次阴性/良性筛查乳房X线检查中的时空变化，以预测病例对照环境中的近期乳腺癌风险。具体而言，LRP-NET是基于临床知识设计的，用于在四次连续乳房X线摄影检查中捕获双侧乳腺组织的成像变化。我们用两个消融研究来评估我们提出的模型，并将其与三个模型/设置进行比较，包括1）没有明确捕获纵向检查的时空变化的“松散”模型，2）LRP-NET，但使用不同的数量（即，1和3）连续检查，以及3）仅使用单个乳房X线照片检查的先前模型。在200例患者的病例对照队列中，每例患者进行4次检查，我们对总共3200张图像进行的实验表明，LRP-NET模型优于比较模型/设置。
Information in digital mammogram images has been shown to be associated with the risk of developing breast cancer. Longitudinal breast cancer screening mammogram examinations may carry spatiotemporal information that can enhance breast cancer risk prediction. No deep learning models have been designed to capture such spatiotemporal information over multiple examinations to predict the risk. In this study, we propose a novel deep learning structure, LRP-NET, to capture the spatiotemporal changes of breast tissue over multiple negative/benign screening mammogram examinations to predict near-term breast cancer risk in a case-control setting. Specifically, LRP-NET is designed based on clinical knowledge to capture the imaging changes of bilateral breast tissue over four sequential mammogram examinations. We evaluate our proposed model with two ablation studies and compare it to three models/settings, including 1) a “loose” model without explicitly capturing the spatiotemporal changes over longitudinal examinations, 2) LRP-NET but using a varying number (i.e., 1 and 3) of sequential examinations, and 3) a previous model that uses only a single mammogram examination. On a case-control cohort of 200 patients, each with four examinations, our experiments on a total of 3200 images show that the LRP-NET model outperforms the compared models/settings.