Walton William C, Kim Seung-Jun, Mullen Lisa A
IEEE Trans Biomed Eng. 2022 Nov;69(11):3538-3550. doi: 10.1109/TBME.2022.3173182. Epub 2022 Oct 19.
Automated registration algorithms for a pair of 2D X-ray mammographic images taken from two standard imaging angles, namely, the craniocaudal (CC) and the mediolateral oblique (MLO) views, are developed.
A fully convolutional neural network, a type of convolutional neural network (CNN), is employed to generate a pixel-level deformation field, which provides a mapping between masses in the two views. Novel distance-based regularization is employed, which contributes significantly to the performance.
The developed techniques are tested using real 2D mammographic images, slices from real 3D mammographic images, and synthetic mammographic images. Architectural variations of the neural network are investigated and the performance is characterized from various aspects including image resolution, breast density, lesion size, lesion subtlety, and lesion Breast Imaging-Reporting and Data System (BI-RADS) category. Our network outperformed the state-of-the-art CNN-based and non-CNN-based registration techniques, and showed robust performance across various tissue/lesion characteristics.
The proposed methods provide a useful automated tool for co-locating lesions between the CC and MLO views even in challenging cases.
Our methods can aid clinicians to establish lesion correspondence quickly and accurately in the dual-view X-ray mammography, improving diagnostic capability.
开发用于一对从两个标准成像角度（即头尾位（CC）和内外斜位（MLO）视图）获取的二维乳腺X线图像的自动配准算法。
采用一种全卷积神经网络（一种卷积神经网络（CNN）类型）来生成像素级变形场，该变形场提供两个视图中肿块之间的映射。采用了基于距离的新型正则化方法，这对性能有显著贡献。
使用真实的二维乳腺X线图像、真实三维乳腺X线图像的切片以及合成乳腺X线图像对所开发的技术进行了测试。研究了神经网络的架构变化，并从图像分辨率、乳腺密度、病变大小、病变细微程度以及病变乳腺影像报告和数据系统（BI-RADS）类别等各个方面对性能进行了表征。我们的网络优于基于CNN和非基于CNN的现有配准技术，并且在各种组织/病变特征方面表现出稳健的性能。
所提出的方法提供了一种有用的自动工具，即使在具有挑战性的情况下也能在CC和MLO视图之间对病变进行共定位。
我们的方法可以帮助临床医生在双视图乳腺X线摄影中快速准确地建立病变对应关系，提高诊断能力。
