Chen Hung-Hsun, Liu Chih-Min, Chang Shih-Lin, Chang Paul Yu-Chun, Chen Wei-Shiang, Pan Yo-Ming, Fang Ssu-Ting, Zhan Shan-Quan, Chuang Chieh-Mao, Lin Yenn-Jiang, Kuo Ling, Wu Mei-Han, Chen Chun-Ku, Chang Ying-Yueh, Shiu Yang-Che, Chen Shih-Ann, Lu Henry Horng-Shing
Center of Teaching and Learning Development, National Chiao Tung University, Taiwan.
Heart Rhythm Center, Division of Cardiology, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan; Faculty of Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan; Institute of Clinical Medicine, National Yang-Ming University, Taipei, Taiwan.
Int J Cardiol. 2020 Oct 1;316:272-278. doi: 10.1016/j.ijcard.2020.03.075. Epub 2020 Apr 11.
Precise segmentation of the left atrium (LA) in computed tomography (CT) images constitutes a crucial preparatory step for catheter ablation in atrial fibrillation (AF). We aim to apply deep convolutional neural networks (DCNNs) to automate the LA detection/segmentation procedure and create three-dimensional (3D) geometries.
Five hundred eighteen patients who underwent procedures for circumferential isolation of four pulmonary veins were enrolled. Cardiac CT images (from 97 patients) were used to construct the LA detection and segmentation models. These images were reviewed by the cardiologists such that images containing the LA were identified/segmented as the ground truth for model training. Two DCNNs which incorporated transfer learning with the architectures of ResNet50/U-Net were trained for image-based LA classification/segmentation. The LA geometry created by the deep learning model was correlated to the outcomes of AF ablation.
The LA detection model achieved an overall 99.0% prediction accuracy, as well as a sensitivity of 99.3% and a specificity of 98.7%. Moreover, the LA segmentation model achieved an intersection over union of 91.42%. The estimated mean LA volume of all the 518 patients studied herein with the deep learning model was 123.3 ± 40.4 ml. The greatest area under the curve with a LA volume of 139 ml yielded a positive predictive value of 85.5% without detectable AF episodes over a period of one year following ablation.
The deep learning provides an efficient and accurate way for automatic contouring and LA volume calculation based on the construction of the 3D LA geometry.
在计算机断层扫描（CT）图像中精确分割左心房（LA）是心房颤动（AF）导管消融的关键准备步骤。我们旨在应用深度卷积神经网络（DCNN）实现LA检测/分割过程的自动化并创建三维（3D）几何模型。
纳入518例行肺静脉环形隔离术的患者。使用心脏CT图像（来自97例患者）构建LA检测和分割模型。心脏病专家对这些图像进行了审查，以便识别/分割包含LA的图像作为模型训练的基准真值。使用两种结合了迁移学习且采用ResNet50/U-Net架构的DCNN进行基于图像的LA分类/分割训练。将深度学习模型创建的LA几何模型与AF消融结果相关联。
LA检测模型的总体预测准确率达到99.0%，灵敏度为99.3%，特异性为98.7%。此外，LA分割模型的交并比达到91.42%。使用深度学习模型对本文研究的所有518例患者估计的平均LA体积为123.3±40.4ml。LA体积为139ml时曲线下面积最大，在消融后一年的时间内无可检测到的AF发作，其阳性预测值为85.5%。
深度学习基于3D LA几何模型的构建，为自动轮廓描绘和LA体积计算提供了一种高效且准确的方法。
