Guangdong Academy Research on VR Industry, Foshan University, Guangdong, People's Republic of China.
Department of Cardiology, Shanghai Chest Hospital, Shanghai JiaoTong University, Shanghai, People's Republic of China.
Am J Physiol Heart Circ Physiol. 2021 Aug 1;321(2):H390-H399. doi: 10.1152/ajpheart.00416.2020. Epub 2021 Jun 25.
Deep learning (DL) has been applied for automatic left ventricle (LV) ejection fraction (EF) measurement, but the diagnostic performance was rarely evaluated for various phenotypes of heart disease. This study aims to evaluate a new DL algorithm for automated LVEF measurement using two-dimensional echocardiography (2DE) images collected from three centers. The impact of three ultrasound machines and three phenotypes of heart diseases on the automatic LVEF measurement was evaluated. Using 36890 frames of 2DE from 340 patients, we developed a DL algorithm based on U-Net (DPS-Net) and the biplane Simpson's method was applied for LVEF calculation. Results showed a high performance in LV segmentation and LVEF measurement across phenotypes and echo systems by using DPS-Net. Good performance was obtained for LV segmentation when DPS-Net was tested on the CAMUS data set (Dice coefficient of 0.932 and 0.928 for ED and ES). Better performance of LV segmentation in study-wise evaluation was observed by comparing the DPS-Net v2 to the EchoNet-dynamic algorithm ( = 0.008). DPS-Net was associated with high correlations and good agreements for the LVEF measurement. High diagnostic performance was obtained that the area under receiver operator characteristic curve was 0.974, 0.948, 0.968, and 0.972 for normal hearts and disease phenotypes including atrial fibrillation, hypertrophic cardiomyopathy, dilated cardiomyopathy, respectively. High performance was obtained by using DPS-Net in LV detection and LVEF measurement for heart failure with several phenotypes. High performance was observed in a large-scale dataset, suggesting that the DPS-Net was highly adaptive across different echocardiographic systems. A new strategy of feature extraction and fusion could enhance the accuracy of automatic LVEF assessment based on multiview 2-D echocardiographic sequences. High diagnostic performance for the determination of heart failure was obtained by using DPS-Net in cases with different phenotypes of heart diseases. High performance for left ventricle segmentation was obtained by using DPS-Net, suggesting the potential for a wider range of application in the interpretation of 2DE images.
深度学习（DL）已被应用于自动左心室（LV）射血分数（EF）测量，但对于各种心脏病表型的诊断性能却很少进行评估。本研究旨在评估一种新的基于二维超声心动图（2DE）图像的深度学习算法，该算法来自三个中心。评估了三种超声机器和三种心脏病表型对自动 LVEF 测量的影响。使用来自 340 名患者的 36890 个 2DE 帧，我们开发了一种基于 U-Net（DPS-Net）的深度学习算法，并应用双平面 Simpson 法进行 LVEF 计算。结果表明，通过使用 DPS-Net，在各种表型和超声系统中，LV 分割和 LVEF 测量的性能都很高。在使用 DPS-Net 对 CAMUS 数据集进行测试时，LV 分割获得了很好的性能（ED 和 ES 的 Dice 系数分别为 0.932 和 0.928）。通过比较 DPS-Net v2 与 EchoNet-dynamic 算法，观察到在研究内评估中 LV 分割的性能更好（ = 0.008）。DPS-Net 与 LVEF 测量具有高度相关性和良好一致性。获得了较高的诊断性能，对于正常心脏和包括房颤、肥厚型心肌病、扩张型心肌病在内的疾病表型，接收器操作特征曲线下面积分别为 0.974、0.948、0.968 和 0.972。对于具有多种表型的心力衰竭，DPS-Net 在 LV 检测和 LVEF 测量中获得了较高的性能。在大规模数据集上获得了较高的性能，表明 DPS-Net 对不同的超声心动图系统具有很强的适应性。基于多视图 2D 超声心动图序列的特征提取和融合的新策略可以提高自动 LVEF 评估的准确性。使用 DPS-Net 在具有不同心脏病表型的病例中，获得了心力衰竭的高诊断性能。通过使用 DPS-Net 获得了较高的 LV 分割性能，这表明它在 2DE 图像解释中有更广泛的应用潜力。
