Department of Radiology and Radiological Science, Division of Cardiovascular Imaging, Medical University of South Carolina, Charleston, SC.
Department of Radiology, University Medical Center Groningen, Groningen, The Netherlands.
J Thorac Imaging. 2020 May;35 Suppl 1:S35-S39. doi: 10.1097/RTI.0000000000000484.
The purpose of this study was to validate the accuracy of an artificial intelligence (AI) prototype application in determining bone mineral density (BMD) from chest computed tomography (CT), as compared with dual-energy x-ray absorptiometry (DEXA).
In this Institutional Review Board-approved study, we analyzed the data of 65 patients (57 female, mean age: 67.4 y) who underwent both DEXA and chest CT (mean time between scans: 1.31 y). From the DEXA studies, T-scores for L1-L4 (lumbar vertebrae 1 to 4) were recorded. Patients were then divided on the basis of their T-scores into normal control, osteopenic, or osteoporotic groups. An AI algorithm based on wavelet features, AdaBoost, and local geometry constraints independently localized thoracic vertebrae from chest CT studies and automatically computed average Hounsfield Unit (HU) values with kVp-dependent spectral correction. The Pearson correlation evaluated the correlation between the T-scores and HU values. Mann-Whitney U test was implemented to compare the HU values of normal control versus osteoporotic patients.
Overall, the DEXA-determined T-scores and AI-derived HU values showed a moderate correlation (r=0.55; P<0.001). This 65-patient population was divided into 3 subgroups on the basis of their T-scores. The mean T-scores for the 3 subgroups (normal control, osteopenic, osteoporotic) were 0.77±1.50, -1.51±0.04, and -3.26±0.59, respectively. The mean DEXA-determined L1-L4 BMD measures were 1.13±0.16, 0.88±0.06, and 0.68±0.06 g/cm, respectively. The mean AI-derived attenuation values were 145±42.5, 136±31.82, and 103±16.28 HU, respectively. Using these AI-derived HU values, a significant difference was found between the normal control patients and osteoporotic group (P=0.045).
Our results show that this AI prototype can successfully determine BMD in moderate correlation with DEXA. Combined with other AI algorithms directed at evaluating cardiac and lung diseases, this prototype may contribute to future comprehensive preventative care based on a single chest CT.
本研究旨在验证一种人工智能（AI）原型应用程序从胸部计算机断层扫描（CT）确定骨密度（BMD）的准确性，与双能 X 射线吸收法（DEXA）进行比较。
在这项经过机构审查委员会批准的研究中，我们分析了 65 名患者（57 名女性，平均年龄：67.4 岁）的数据，这些患者均同时接受了 DEXA 和胸部 CT 检查（两次扫描的平均时间间隔为 1.31 年）。从 DEXA 研究中，记录了腰椎 1 到 4 （L1-L4）的 T 评分。然后，根据 T 评分将患者分为正常对照组、骨质疏松组或骨质疏松症组。一种基于小波特征、AdaBoost 和局部几何约束的 AI 算法，从胸部 CT 研究中独立定位胸椎，并自动计算平均亨氏单位（HU）值，同时进行与千伏相关的光谱校正。Pearson 相关性评估了 T 评分与 HU 值之间的相关性。采用 Mann-Whitney U 检验比较正常对照组与骨质疏松症组的 HU 值。
总体而言，DEXA 确定的 T 评分和 AI 推导的 HU 值显示出中度相关性（r=0.55；P<0.001）。基于 T 评分，将这 65 名患者分为 3 个亚组。3 个亚组（正常对照组、骨质疏松组、骨质疏松症组）的平均 T 评分分别为 0.77±1.50、-1.51±0.04 和-3.26±0.59。平均 DEXA 测定的 L1-L4 BMD 测量值分别为 1.13±0.16、0.88±0.06 和 0.68±0.06 g/cm。平均 AI 推导的衰减值分别为 145±42.5、136±31.82 和 103±16.28 HU。使用这些 AI 推导的 HU 值，正常对照组和骨质疏松症组之间存在显著差异（P=0.045）。
我们的结果表明，该 AI 原型可以与 DEXA 成功地确定骨密度，相关性中等。结合其他针对评估心脏和肺部疾病的 AI 算法，该原型可能有助于未来基于单次胸部 CT 的综合预防性护理。
