深度学习增强的100 kV低电压胸部CT机会性骨质疏松筛查:骨密度测量与辐射剂量降低的新途径
本研究纳入1167例同时行100 kV低电压胸部CT和120 kV腰椎CT的患者,用深度学习实现椎体自动分割、ROI提取和骨密度计算,以120 kV QCT为参考。结果显示100 kV CNN的骨密度与120 kV QCT相关性更强、一致性更好,诊断骨质疏松和低骨密度的AUC达1.000,并在不同厂商和医院的多种扫描仪上表现良好。
要点
- 100 kV低电压胸部CT联合深度学习可全自动完成T12-L2椎体骨密度测量。
- 100 kV CNN骨密度与120 kV QCT一致性优于100 kV QCT。
- 100 kV CNN诊断骨质疏松和低骨密度AUC达1.000。
核心结论
- 深度学习方法在肺癌筛查的100 kV低电压胸部CT上实现高精度全自动骨质疏松筛查,并在不同厂商和医院的多款扫描仪上表现良好。
研究设计与样本
- 研究纳入2022年10月至2024年8月同时行100 kV低电压胸部CT和120 kV腰椎CT的1167例患者,分为训练集495例、验证集169例和三个测试集245例、128例、130例;以120 kV QCT骨密度值为参考。
关键影像表现/征象
- 深度学习框架含四个卷积神经网络:3D VB-Net和SCN用于椎体自动分割与ROI提取,DenseNet和ResNet用于目标椎体T12-L2的骨密度计算。
诊断标准与关键数值
- 三个测试集中,120 kV QCT与100 kV CNN骨密度的线性回归R²=0.970-0.994(P<.001),Bland-Altman平均误差为-2.24至1.52 mg/cm³。
- 三个测试集中,120 kV QCT与100 kV QCT骨密度的线性回归R²=0.968-0.986(P<.001),Bland-Altman平均误差为2.72至3.06 mg/cm³。
- 100 kV CNN和100 kV QCT从正常骨密度中检测骨质疏松的AUC分别为1.000和0.999-1.000。
- 100 kV CNN和100 kV QCT从正常骨密度中检测低骨密度的AUC分别为1.000和1.000。
临床应用与阅片要点
- 可利用100 kV低电压胸部CT进行机会性骨质疏松筛查,深度学习自动测量T12-L2椎体骨密度。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
RATIONALE AND OBJECTIVES: To explore the feasibility and accuracy of a deep learning (DL) method for fully automated vertebral body (VB) segmentation, region of interest (ROI) extraction, and bone mineral density (BMD) calculation using 100kV low-voltage chest CT performed for lung cancer screening across various scanners from different manufacturers and hospitals.
MATERIALS AND METHODS: This study included 1167 patients who underwent 100 kV low-voltage chest and 120 kV lumbar CT from October 2022 to August 2024. Patients were divided into a training set (495 patients), a validation set (169 patients), and three test sets (245, 128, and 130 patients). The DL framework comprised four convolutional neural networks (CNNs): 3D VB-Net and SCN for automated VB segmentation and ROI extraction, and DenseNet and ResNet for BMD calculation of target VBs (T12-L2). The BMD values of 120 kV QCT were identified as reference data. Linear regression and BlandAltman analyses were used to compare the BMD values between 120 kV QCT and 100 kV CNNs and 100 kV QCT. Receiver operating characteristic curve analysis was used to evaluate the diagnostic performance of 100 kV CNNs and 100 kV QCT for osteoporosis and low BMD from normal BMD.
RESULTS: For three test sets, linear regression and BlandAltman analyses revealed a stronger correlation (R2 = 0.970-0.994 and 0.968-0.986, P < .001) and better agreement (mean error, -2.24 to 1.52 and 2.72 to 3.06 mg/cm3) for the BMD between the 120 kV QCT and 100 kV CNNs than between the 120 kV and 100 kV QCT. The areas under the curve of the 100 kV CNNs and 100 kV QCT were 1.000 and 0.999-1.000, and 1.000 and 1.000 for detecting osteoporosis and low BMD from normal BMD, respectively.
CONCLUSION: The DL method achieved high accuracy for fully automated osteoporosis screening in 100 kV low-voltage chest CT scans obtained for lung cancer screening and performed well on various scanners from different manufacturers and hospitals.
引用
- Li Y, Ye K, Liu S, et al. Deep Learning-Enhanced Opportunistic Osteoporosis Screening in 100 kV Low-Voltage Chest CT: A Novel Way Toward Bone Mineral Density Measurement and Radiation Dose Reduction. Acad Radiol. 2025. PMID: 40835570. doi:10.1016/j.acra.2025.07.060
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。