Shiomi Takuma, Nakayama Ryohei, Hizukuri Akiyoshi, Takafuji Masafumi, Ishida Masaki, Sakuma Hajime
Graduate School of Science and Engineering, Ritsumeikan University, 1-1-1 Noji-Higashi, Kusatsu, Shiga, 525-8577, Japan.
Department of Radiology, Mie University School of Medicine, 2-174 Edobashi, Tsu, Mie, 514-8507, Japan.
Radiol Phys Technol. 2025 Mar;18(1):219-226. doi: 10.1007/s12194-024-00875-x. Epub 2024 Dec 31.
This study aims to develop a computerized classification method for significant coronary artery stenosis on whole-heart coronary magnetic resonance angiography (WHCMRA) images using a 3D convolutional neural network (3D-CNN) with attention mechanisms. The dataset included 951 segments from WHCMRA images of 75 patients who underwent both WHCMRA and invasive coronary angiography (ICA). Forty-two segments with significant stenosis (luminal diameter reduction 75%) on ICA were annotated on WHCMRA images by an experienced radiologist, whereas 909 segments without it were annotated at representative sites. Volumes of interest (VOIs) of 21 × 21 × 21 voxels centered on annotated points were extracted. The network comprises two feature extractors, two attention mechanisms (for the coronary artery and annotated points), and a classifier. The feature extractors first extracted the feature maps from the VOI. The two attention mechanisms weighted the feature maps of the coronary artery and those the neighborhood of the annotated point, respectively. The classifier finally classified the VOIs into those with and without significant coronary artery stenosis. Using fivefold cross-validation, the classification accuracy, sensitivity, specificity, and AUROC (area under the receiver operating characteristic curve) were 0.875, 0.905, 0.873, and 0.944, respectively. The proposed method showed high classification performance for significant coronary artery stenosis and appears to have a substantial impact on the interpretation of WHCMRA images.
本研究旨在利用带有注意力机制的三维卷积神经网络（3D-CNN），开发一种用于全心脏冠状动脉磁共振血管造影（WHCMRA）图像中显著冠状动脉狭窄的计算机化分类方法。该数据集包括75例接受了WHCMRA和有创冠状动脉造影（ICA）检查的患者的WHCMRA图像中的951个节段。由一位经验丰富的放射科医生在WHCMRA图像上标注了ICA上有显著狭窄（管腔直径缩小75%）的42个节段，而909个无显著狭窄的节段则在代表性部位进行了标注。提取了以标注点为中心的21×21×21体素的感兴趣体积（VOI）。该网络由两个特征提取器、两个注意力机制（分别针对冠状动脉和标注点）和一个分类器组成。特征提取器首先从VOI中提取特征图。两个注意力机制分别对冠状动脉的特征图和标注点附近的特征图进行加权。分类器最终将VOI分为有显著冠状动脉狭窄和无显著冠状动脉狭窄两类。使用五折交叉验证，分类准确率、灵敏度、特异性和ROC曲线下面积（AUROC）分别为0.875、0.905、0.873和0.944。所提出的方法对显著冠状动脉狭窄显示出较高的分类性能，并且似乎对WHCMRA图像的解读有重大影响。
