Department of Biomedical Engineering, King's College London, Lambeth Palace Rd, London SE1 7EU, UK.
University/BHF Centre for Cardiovascular Science, University of Edinburgh, 47 Little France Crescent, Edinburgh EH16 4TJ, UK.
Eur Heart J Cardiovasc Imaging. 2023 Apr 24;24(5):607-615. doi: 10.1093/ehjci/jead010.
Left atrial volume is commonly estimated using the bi-plane area-length method from two-chamber (2CH) and four-chamber (4CH) long axes views. However, this can be inaccurate due to a violation of geometric assumptions. We aimed to develop a deep learning neural network to infer 3D left atrial shape, volume and surface area from 2CH and 4CH views.
A 3D UNet was trained and tested using 2CH and 4CH segmentations generated from 3D coronary computed tomography angiography (CCTA) segmentations (n = 1700, with 1400/100/200 cases for training/validating/testing). An independent test dataset from another institution was also evaluated, using cardiac magnetic resonance (CMR) 2CH and 4CH segmentations as input and 3D CCTA segmentations as the ground truth (n = 20). For the 200 test cases generated from CCTA, the network achieved a mean Dice score value of 93.7%, showing excellent 3D shape reconstruction from two views compared with the 3D segmentation Dice of 97.4%. The network also showed significantly lower mean absolute error values of 3.5 mL/4.9 cm2 for LA volume/surface area respectively compared to the area-length method errors of 13.0 mL/34.1 cm2 respectively (P < 0.05 for both). For the independent CMR test set, the network achieved accurate 3D shape estimation (mean Dice score value of 87.4%), and a mean absolute error values of 6.0 mL/5.7 cm2 for left atrial volume/surface area respectively, significantly less than the area-length method errors of 14.2 mL/19.3 cm2 respectively (P < 0.05 for both).
Compared to the bi-plane area-length method, the network showed higher accuracy and robustness for both volume and surface area.
左心房容积通常通过双平面（2CH）和四腔（4CH）长轴视图的平面面积长度法进行估计。然而，由于违反了几何假设，这可能会不准确。我们旨在开发一种深度学习神经网络，从 2CH 和 4CH 视图推断 3D 左心房形状、体积和表面积。
使用从 3D 冠状动脉计算机断层扫描血管造影（CCTA）分割中生成的 2CH 和 4CH 分割对 3D UNet 进行了训练和测试（n = 1700，训练/验证/测试分别为 1400/100/200 例）。还评估了来自另一个机构的独立测试数据集，使用心脏磁共振（CMR）2CH 和 4CH 分割作为输入，3D CCTA 分割作为ground truth（n = 20）。对于从 CCTA 生成的 200 个测试病例，网络实现了 93.7%的平均骰子得分值，与 3D 分割的 97.4%相比，从两个视图显示出出色的 3D 形状重建。与平面面积长度法的误差分别为 13.0 mL/34.1 cm2 相比，网络还显示出明显更低的左心房容积/表面积的平均绝对误差值，分别为 3.5 mL/4.9 cm2（均 P < 0.05）。对于独立的 CMR 测试集，网络实现了准确的 3D 形状估计（平均骰子得分值为 87.4%），左心房容积/表面积的平均绝对误差值分别为 6.0 mL/5.7 cm2，明显小于平面面积长度法的误差，分别为 14.2 mL/19.3 cm2（均 P < 0.05）。
与双平面面积长度法相比，该网络在体积和表面积方面均表现出更高的准确性和鲁棒性。
