Department of Radiology, The First Affiliated Hospital of Xi'an Jiaotong University, Xi'an, Shanxi 710061, China; Department of Radiology, Gansu Provincial Hospital of Traditional Chinese Medicine, Lanzhou, Gansu 730050, China.
Gansu University of Traditional Chinese Medicine, Lanzhou, Gansu 730000, China.
Eur J Radiol. 2022 Jan;146:110071. doi: 10.1016/j.ejrad.2021.110071. Epub 2021 Nov 28.
To develop a deep learning-based model for measuring automatic lumbosacral anatomical parameters from lateral lumbar radiographs and compare its performance to that of attending-level radiologists.
A total of 1791 lateral lumbar radiographs were collected through the PACS system and used to develop the deep learning-based model. Landmarks for the four used parameters, including the lumbosacral lordosis angle (LSLA), lumbosacral angle (LSA), sacral horizontal angle (SHA), and sacral inclination angle (SIA), were identified and automatically labeled by the model. At the same time, the measurement results were obtained through landmarks on the test set compared to manual measurements as the reference standard. Statistical analyses of the Percentage of Correct Key Points (PCK), intra-class correlation coefficient (ICC), Pearson correlation coefficient, mean absolute error (MAE), root mean square error (RMSE), and Bland-Altman plots were performed to evaluate the performance of the model.
The mean differences between the reference standard and the model for LSLA, LSA, SHA, and SIA, were 0.39°, 0.09°, 0.13°, and 0.12°, respectively. A strong correlation and consistency between the four parameters were found between the model and reference standard (ICC = 0.92-0.98, r = 0.92-0.97, MAE = 1.35-1.84, RMSE = 1.82-2.51), while with statistically significant difference for LSLA (p = 0.02).
The presented model revealed clinically equivalent measurements in terms of accuracy, while superior measurements were obtained in terms of cost-effectiveness, reliability, and reproducibility. The model may help clinicians improve their understanding and evaluation of lumbar diseases and LBP from a quantitative perspective in practical work. (ChiCTR2100048250).
开发一种基于深度学习的方法，从侧位腰椎 X 光片中测量自动腰骶解剖参数，并将其性能与放射科医师的水平进行比较。
通过 PACS 系统共收集 1791 例侧位腰椎 X 光片用于开发基于深度学习的模型。该模型自动识别并标记了四个使用参数（腰骶角、腰骶角、骶骨水平角和骶骨倾斜角）的标志点。同时，通过与手动测量结果作为参考标准，比较测试集上的测量结果。采用关键点正确百分比（PCK）、组内相关系数（ICC）、皮尔逊相关系数、平均绝对误差（MAE）、均方根误差（RMSE）和 Bland-Altman 图对模型性能进行评估。
参考标准与模型之间 LSLA、LSA、SHA 和 SIA 的平均差异分别为 0.39°、0.09°、0.13°和 0.12°。模型与参考标准之间四个参数均具有较强的相关性和一致性（ICC=0.92-0.98，r=0.92-0.97，MAE=1.35-1.84，RMSE=1.82-2.51），但 LSLA 存在统计学差异（p=0.02）。
该模型在准确性方面具有临床等效的测量结果，在成本效益、可靠性和可重复性方面具有更优的测量结果。该模型可能有助于临床医生从定量角度提高对腰椎疾病和下腰痛的理解和评估。（ChiCTR2100048250）
