Artificial intelligence X-ray measurement technology of anatomical parameters related to lumbosacral stability. Issue 146 (January 2022)

• Record Type: Journal Article
• Authors: Zhou, Sheng; Yao, Hongyan; Ma, Chunyu; Chen, Xiaofei; Wang, Wenqi; Ji, Hongli; He, Linyang; Luo, Mengyan; Guo, Youmin
• Is Part Of: European journal of radiology. Issue 146(2022)
• Publication Date: 2022-01
• Subjects: Low back pain -- Lateral lumbar radiograph -- Lumbosacral parameters -- Deep learning -- Automatic measurement; AI Artificial intelligence -- BiFPN Bi-directional feature pyramid network -- CNN Convolutional neural network -- CPN Cascaded Pyramid Network -- DL Deep learning -- FCN Fully Connected Network -- ICC Intra-class correlation coefficient -- LSLA Lumbosacral lordosis angle -- LSA Lumbosacral angle -- LBP Low back pain -- LoA Limit of agreement -- MAE Mean absolute error -- PACS Picture Archiving and Communication System -- PCK Percentage of Correct Key Points -- RMSE Root mean square error -- SHA Sacral horizontal angle -- SIA Sacral inclination angle -- SD standard deviation; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2021.110071
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: Accurate measurement of lumbosacral anatomical parameters is significant for quantitatively evaluating lumbar diseases. Deep learning algorithm can be used to develop an automatic measurement model based on lateral lumbar radiography. The performance of the model is superior or equal to a radiologist in measuring the LSLA, LSA, SIA, and SHA. Abstract: Purpose: 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. Methods: 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. Results: 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). Conclusions: 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).