Wang Huayu, Hu Yixin, Lu Yao, Zhou Jianhua, Guo Yongze
School of Computer Science and Engineering, Sun Yat-Sen University, Guangzhou, P.R. China.
Department of Ultrasound, Sun Yat-Sen University Cancer Center, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Guangzhou, P.R. China.
Med Phys. 2022 May;49(5):3314-3324. doi: 10.1002/mp.15590. Epub 2022 Mar 15.
The Breast Imaging-Reporting and Data System (BI-RADS) for ultrasound imaging provides a widely used reporting schema for breast imaging. Previous studies have shown that in ultrasound imaging, 90% of BI-RADS 4A tumors are benign lesions after biopsies. Unnecessary biopsy procedures can be avoided by accurate classification of BI-RADS 4A tumors. However, the classification task is challenging and has not been fully investigated by existing studies. For benign and malignant tumors of BI-RADS 4A, the appearances of intra-class tumors are highly variable, the characteristics of inter-class tumors is overall-similar. Discriminative features need to be found to improve classification accuracy of BI-RADS 4A tumors.
In this study, we designed the network using the clinical features of BI-RADS 4A tumors to improve the discrimination ability of network. The boundary information is embedded into the input of the network using the uncertainty. A fine-grained data augmentation method is used to find discriminative features in tumor information embedded with boundary information. Two mathematical methods, voting-based and variance-based, are used to define the uncertainty of boundary, and the differences of these two definitions are compared in a classification network.
The dataset we used to evaluate our method had 1155 2D grayscale images. Each image represented a unique BI-RADS 4A tumor. Among them, 248 tumors were proven to be malignant by biopsy, and the remaining 907 were benign. A weakly supervised data augmentation network (WS-DAN) was used as the backbone classification network, which showed competitive performance in finding discriminative features. Using the auxiliary input of the uncertain boundaries defined by the voting method, the area under the curve (AUC) value of our method was 0.8347 (sensitivity = 0.7774, specificity = 0.7459). The AUC value of the variance-based uncertainty was 0.7789. The voting-based uncertainty was higher than the baseline (AUC = 0.803), which only inputs the original image. Compared with the classic classification network, our method had a significant effect improvement (p < 0.01).
Using the uncertain boundaries defined by the voting methods as auxiliary information, we obtained a better performance in the classification of BI-RADS 4A ultrasound images, while variance-based uncertain boundaries had no effect on improving classification performance. Additionally, fine-grained network helped find discriminative features comparing with the commonly used classification networks.
超声成像的乳腺影像报告和数据系统（BI-RADS）为乳腺成像提供了一种广泛使用的报告模式。先前的研究表明，在超声成像中，90%的BI-RADS 4A类肿瘤活检后为良性病变。通过对BI-RADS 4A类肿瘤进行准确分类，可以避免不必要的活检程序。然而，分类任务具有挑战性，现有研究尚未对其进行充分研究。对于BI-RADS 4A类的良性和恶性肿瘤，类内肿瘤的表现高度可变，类间肿瘤的特征总体相似。需要找到判别特征以提高BI-RADS 4A类肿瘤的分类准确性。
在本研究中，我们利用BI-RADS 4A类肿瘤的临床特征设计网络，以提高网络的判别能力。利用不确定性将边界信息嵌入到网络输入中。采用细粒度数据增强方法在嵌入边界信息的肿瘤信息中寻找判别特征。使用基于投票和基于方差的两种数学方法定义边界的不确定性，并在分类网络中比较这两种定义的差异。
我们用于评估方法的数据集有1155张二维灰度图像。每张图像代表一个独特的BI-RADS 4A类肿瘤。其中，248个肿瘤经活检证实为恶性，其余907个为良性。使用弱监督数据增强网络（WS-DAN）作为主干分类网络，其在寻找判别特征方面表现出有竞争力的性能。使用投票方法定义的不确定边界的辅助输入，我们方法的曲线下面积（AUC）值为0.8347（灵敏度=0.7774，特异性=0.7459）。基于方差的不确定性的AUC值为0.7789。基于投票产生的不确定性高于仅输入原始图像的基线（AUC=0.803）。与经典分类网络相比，我们的方法有显著的效果提升（p<0.01）。
将投票方法定义的不确定边界用作辅助信息，我们在BI-RADS 4A类超声图像分类中获得了更好的性能，而基于方差的不确定边界对提高分类性能没有作用。此外，与常用的分类网络相比，细粒度网络有助于找到判别特征。
