Graduate School of Environment and Information Sciences, Yokohama National University, 79-7 Tokiwadai, Hodogaya-ku, Yokohama, 240-8501, Japan.
Department of Advanced Biomedical Imaging and Informatics, St. Marianna University School of Medicine, 2-16-1 Sugao, Miyamae-ku, Kawasaki, Kanagawa, 216-8511, Japan.
Radiol Phys Technol. 2022 Dec;15(4):340-348. doi: 10.1007/s12194-022-00673-3. Epub 2022 Aug 21.
The purpose of this study was to verify the efficacy of generative contribution mapping (GCM), an explainable deep learning model for images, in classifying the presence or absence of calcifications on mammography. The learning dataset consisted of 303 full-field digital mammography (FFDM) images labeled with microcalcifications obtained from the public INbreast database without extremely dense images. FFDM images were divided into calcification and non-calcification patch images using a sliding window method with 25% overlap. The patch images of the mediolateral oblique (MLO) and craniocaudal (CC) views were divided into a training set of 70%, a validation set of 10%, and a testing set of 20%. The classification performance of GCM classifiers was evaluated and compared with that of EfficientNet classifiers. Visualization maps of GCM highlighted regions of interest more clearly than EfficientNet's gradient-weighted class activation maps. The results showed that GCM classifiers yielded an accuracy of 0.92 (CC), 0.91 (MLO), and an area under the receiver operating characteristic curve of 0.92 (CC), 0.94 (MLO). In conclusion, GCM could accurately classify the presence or absence of calcifications on mammograms and explain intuitively reasonable grounds for their classification with visualization maps highlighting regions of interest.
本研究旨在验证生成式贡献映射（GCM）的有效性，这是一种用于图像的可解释深度学习模型，用于对乳腺 X 光片中是否存在钙化进行分类。学习数据集由来自公共 INbreast 数据库的带有微钙化的 303 张全视野数字化乳腺 X 光摄影（FFDM）图像组成，不包括非常致密的图像。使用具有 25%重叠的滑动窗口方法将 FFDM 图像分为钙化和非钙化斑块图像。MLO 和 CC 视图的斑块图像分为训练集的 70%、验证集的 10%和测试集的 20%。评估和比较了 GCM 分类器和 EfficientNet 分类器的分类性能。GCM 分类器的可视化映射比 EfficientNet 的梯度加权类激活映射更清晰地突出了感兴趣的区域。结果表明，GCM 分类器在 CC 视图中的准确率为 0.92，在 MLO 视图中的准确率为 0.91，在 CC 视图中的接收者操作特征曲线下面积为 0.92，在 MLO 视图中的面积为 0.94。总之，GCM 可以准确地对乳腺 X 光片中是否存在钙化进行分类，并通过可视化映射直观地解释分类的合理依据，突出感兴趣的区域。
