Department of Radiology, The Institute of Medical Science, The University of Tokyo, 4-6-1 Shirokanedai, Minato-ku, Tokyo, 108-8639, Japan.
Department of Radiology, Graduate School of Medical Sciences, International University of Health and Welfare, 537-3 Iguchi, Nasushiobara, Tochigi, 329-2763, Japan.
Eur Radiol. 2020 Jun;30(6):3549-3557. doi: 10.1007/s00330-020-06677-0. Epub 2020 Feb 14.
To investigate whether a deep learning model can predict the bone mineral density (BMD) of lumbar vertebrae from unenhanced abdominal computed tomography (CT) images.
In this Institutional Review Board-approved retrospective study, patients who received both unenhanced CT examinations and dual-energy X-ray absorptiometry (DXA) of the lumbar vertebrae, in two institutions (1 and 2), were included. Supervised deep learning was employed to obtain a convolutional neural network (CNN) model using axial CT images, including the lumbar vertebrae as input data and BMD values obtained with DXA as reference data. For this purpose, 1665 CT images from 183 patients in institution 1, which were augmented to 99,900 (= 1665 × 60) images (noise adding, parallel shift and rotation were performed), were used. Internal (by using data of 45 other patients in institution 1) and external validations (by using data of 50 patients in institution 2) were performed to evaluate the performance of the trained CNN model. Correlations and diagnostic performances were evaluated with Pearson's correlation coefficient (r) and area under the receiver operating characteristic curve (AUC), respectively.
The estimated BMD values, according to the CNN model (BMD), were significantly correlated with the BMD values obtained with DXA (r = 0.852 (p < 0.001) and 0.840 (p < 0.001) for the internal and external validation datasets, respectively). Using BMD, osteoporosis was diagnosed with AUCs of 0.965 and 0.970 for the internal and external validation datasets, respectively.
Using deep learning, the BMD of lumbar vertebrae could be predicted from unenhanced abdominal CT images.
• By applying a deep learning technique, the bone mineral density (BMD) of lumbar vertebrae can be estimated from unenhanced abdominal CT images. • A strong correlation was observed between the estimated BMD and the BMD obtained with DXA. • By using the estimated BMD, osteoporosis could be diagnosed with high performance.
研究深度学习模型是否可以从未增强腹部 CT 图像预测腰椎骨密度（BMD）。
在这项经机构审查委员会批准的回顾性研究中，纳入了在两个机构（1 和 2）接受未增强 CT 检查和腰椎双能 X 射线吸收法（DXA）的患者。使用轴向 CT 图像（包括腰椎）作为输入数据，以 DXA 获得的 BMD 值作为参考数据，通过监督深度学习获得卷积神经网络（CNN）模型。为此，使用了来自机构 1 的 183 名患者的 1665 张 CT 图像（进行了噪声添加、平行移位和旋转等增强处理，得到 99900 张图像（1665×60））。进行了内部验证（使用机构 1 的另外 45 名患者的数据）和外部验证（使用机构 2 的 50 名患者的数据），以评估训练后的 CNN 模型的性能。使用 Pearson 相关系数（r）和受试者工作特征曲线下的面积（AUC）分别评估相关性和诊断性能。
根据 CNN 模型（BMD）估计的 BMD 值与通过 DXA 获得的 BMD 值显著相关（内部验证数据集的 r 值分别为 0.852（p<0.001）和 0.840（p<0.001），外部验证数据集的 r 值分别为 0.852（p<0.001）和 0.840（p<0.001））。使用 BMD，内部和外部验证数据集的骨质疏松症诊断 AUC 分别为 0.965 和 0.970。
使用深度学习，可以从未增强的腹部 CT 图像预测腰椎 BMD。
• 通过应用深度学习技术，可以从未增强的腹部 CT 图像估计腰椎骨密度（BMD）。• 观察到估计的 BMD 与 DXA 获得的 BMD 之间存在很强的相关性。• 使用估计的 BMD，可以进行高准确性的骨质疏松症诊断。
