Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 May 8;1(3):180091.
doi: 10.1148/ryai.2019180091.

Fully Automated Diagnosis of Anterior Cruciate Ligament Tears on Knee MR Images by Using Deep Learning

Fang Liu  1 , Bochen Guan  1 , Zhaoye Zhou  1 , Alexey Samsonov  1 , Humberto Rosas  1 , Kevin Lian  1 , Ruchi Sharma  1 , Andrew Kanarek  1 , John Kim  1 , Ali Guermazi  1 , Richard Kijowski  1
Affiliations

Fully Automated Diagnosis of Anterior Cruciate Ligament Tears on Knee MR Images by Using Deep Learning

Fang Liu et al. Radiol Artif Intell. .

Abstract

Purpose: To investigate the feasibility of using a deep learning-based approach to detect an anterior cruciate ligament (ACL) tear within the knee joint at MRI by using arthroscopy as the reference standard.

Materials and methods: A fully automated deep learning-based diagnosis system was developed by using two deep convolutional neural networks (CNNs) to isolate the ACL on MR images followed by a classification CNN to detect structural abnormalities within the isolated ligament. With institutional review board approval, sagittal proton density-weighted and fat-suppressed T2-weighted fast spin-echo MR images of the knee in 175 subjects with a full-thickness ACL tear (98 male subjects and 77 female subjects; average age, 27.5 years) and 175 subjects with an intact ACL (100 male subjects and 75 female subjects; average age, 39.4 years) were retrospectively analyzed by using the deep learning approach. Sensitivity and specificity of the ACL tear detection system and five clinical radiologists for detecting an ACL tear were determined by using arthroscopic results as the reference standard. Receiver operating characteristic (ROC) analysis and two-sided exact binomial tests were used to further assess diagnostic performance.

Results: The sensitivity and specificity of the ACL tear detection system at the optimal threshold were 0.96 and 0.96, respectively. In comparison, the sensitivity of the clinical radiologists ranged between 0.96 and 0.98, while the specificity ranged between 0.90 and 0.98. There was no statistically significant difference in diagnostic performance between the ACL tear detection system and clinical radiologists at P < .05. The area under the ROC curve for the ACL tear detection system was 0.98, indicating high overall diagnostic accuracy.

Conclusion: There was no significant difference between the diagnostic performance of the ACL tear detection system and clinical radiologists for determining the presence or absence of an ACL tear at MRI.© RSNA, 2019Supplemental material is available for this article.

PubMed Disclaimer

Conflict of interest statement

Disclosures of Conflicts of Interest: F.L. disclosed no relevant relationships. B.G. disclosed no relevant relationships. Z.Z. disclosed no relevant relationships. A.S. disclosed no relevant relationships. H.R. disclosed no relevant relationships. K.L. disclosed no relevant relationships. R.S. disclosed no relevant relationships. A.K. disclosed no relevant relationships. J.K. disclosed no relevant relationships. A.G. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: is a consultant for Pfizer, AstraZeneca, Roche, Galapagos, MerckSerono, and TissueGene; holds stock or stock options in BICL. Other relationships: disclosed no relevant relationships. R.K. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: is a consultant for Boston Imaging Core Lab; institution has grants or grants pending with GE Healthcare. Other relationships: disclosed no relevant relationships.

Figures

Figure 1:
Figure 1:
Illustration of the convolutional neural network (CNN) architecture for the deep learning–based anterior cruciate ligament (ACL) tear detection system. The proposed method consisted of three separate CNNs connected in a cascaded fashion to create a fully automated image processing pipeline. BN = batch normalization, Conv2D = 2D convolution, ReLU = rectified linear activation, 2D = two-dimensional.
Figure 2a:
Figure 2a:
Images in a 28-year-old man with a surgically confirmed anterior cruciate ligament (ACL) tear that the ACL tear detection system interpreted as present. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show disruption of fibers and increased signal within the ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrow).
Figure 2b:
Figure 2b:
Images in a 28-year-old man with a surgically confirmed anterior cruciate ligament (ACL) tear that the ACL tear detection system interpreted as present. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show disruption of fibers and increased signal within the ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrow).
Figure 2c:
Figure 2c:
Images in a 28-year-old man with a surgically confirmed anterior cruciate ligament (ACL) tear that the ACL tear detection system interpreted as present. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show disruption of fibers and increased signal within the ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrow).
Figure 2d:
Figure 2d:
Images in a 28-year-old man with a surgically confirmed anterior cruciate ligament (ACL) tear that the ACL tear detection system interpreted as present. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show disruption of fibers and increased signal within the ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrow).
Figure 3a:
Figure 3a:
Images in a 26-year-old man with a surgically confirmed anterior cruciate ligament (ACL) tear that the ACL tear detection system interpreted as present. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show disruption of fibers and increased signal within the ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrows).
Figure 3b:
Figure 3b:
Images in a 26-year-old man with a surgically confirmed anterior cruciate ligament (ACL) tear that the ACL tear detection system interpreted as present. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show disruption of fibers and increased signal within the ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrows).
Figure 3c:
Figure 3c:
Images in a 26-year-old man with a surgically confirmed anterior cruciate ligament (ACL) tear that the ACL tear detection system interpreted as present. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show disruption of fibers and increased signal within the ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrows).
Figure 3d:
Figure 3d:
Images in a 26-year-old man with a surgically confirmed anterior cruciate ligament (ACL) tear that the ACL tear detection system interpreted as present. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show disruption of fibers and increased signal within the ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrows).
Figure 4a:
Figure 4a:
Cropped sagittal MR images in a 20-year-old man with a surgically confirmed anterior cruciate ligament (ACL) tear that the ACL tear detection system interpreted as an intact ACL. (a) Proton density–weighted fast spin-echo image and (b) fat-suppressed T2-weighted fast spin-echo image of the knee analyzed by the classification convolutional neural network show fiber disruption and increased signal within the torn ACL (arrow).
Figure 4b:
Figure 4b:
Cropped sagittal MR images in a 20-year-old man with a surgically confirmed anterior cruciate ligament (ACL) tear that the ACL tear detection system interpreted as an intact ACL. (a) Proton density–weighted fast spin-echo image and (b) fat-suppressed T2-weighted fast spin-echo image of the knee analyzed by the classification convolutional neural network show fiber disruption and increased signal within the torn ACL (arrow).
Figure 5a:
Figure 5a:
Images in a 21-year-old man with a surgically confirmed intact anterior cruciate ligament (ACL) that the ACL tear detection system interpreted as an ACL tear. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show continuous fibers and normal signal within the intact ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrows).
Figure 5b:
Figure 5b:
Images in a 21-year-old man with a surgically confirmed intact anterior cruciate ligament (ACL) that the ACL tear detection system interpreted as an ACL tear. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show continuous fibers and normal signal within the intact ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrows).
Figure 5c:
Figure 5c:
Images in a 21-year-old man with a surgically confirmed intact anterior cruciate ligament (ACL) that the ACL tear detection system interpreted as an ACL tear. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show continuous fibers and normal signal within the intact ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrows).
Figure 5d:
Figure 5d:
Images in a 21-year-old man with a surgically confirmed intact anterior cruciate ligament (ACL) that the ACL tear detection system interpreted as an ACL tear. (a) Cropped sagittal proton density–weighted fast spin-echo MR image and (b) cropped sagittal fat-suppressed T2-weighted fast spin-echo MR image of the knee analyzed by the classification convolutional neural network show continuous fibers and normal signal within the intact ACL (arrow). (c) Probability map for the proton density–weighted fast spin-echo image and (d) probability map for the sagittal fat-suppressed T2-weighted fast spin-echo image show the high-probability areas in the ligament on which the machine based its interpretation of an ACL tear (arrows).
Figure 6:
Figure 6:
Receiver operating characteristic (ROC) curves describing the diagnostic performance of the proposed anterior cruciate ligament (ACL) tear detection system for determining the presence or absence of a surgically confirmed ACL tear. The area under the ROC curve for the machine was 0.98, indicating high overall diagnostic accuracy. Sensitivity and specificity for the musculoskeletal radiologist, musculoskeletal radiology fellow, radiology residents, and machine at the optimal threshold of the Youden index are also plotted. Note that the sensitivity and specificity of the clinical radiologists are in close proximity to the ROC curve of the ACL tear detection system.
See this image and copyright information in PMC

References

    1. Liu F, Jang H, Kijowski R, Bradshaw T, McMillan AB. Deep learning MR imaging-based attenuation correction for PET/MR imaging. Radiology 2018;286(2):676–684. - PMC - PubMed
    1. Zhu B, Liu JZ, Cauley SF, Rosen BR, Rosen MS. Image reconstruction by domain-transform manifold learning. Nature 2018;555(7697):487–492. - PubMed
    1. Chen F, Taviani V, Malkiel I, et al. Variable-density single-shot fast spin-echo MRI with deep learning reconstruction by using variational networks. Radiology 2018;289(2):366–373. - PMC - PubMed
    1. Liu F, Zhou Z, Jang H, Samsonov A, Zhao G, Kijowski R. Deep convolutional neural network and 3D deformable approach for tissue segmentation in musculoskeletal magnetic resonance imaging. Magn Reson Med 2018;79(4):2379–2391. - PMC - PubMed
    1. Norman B, Pedoia V, Majumdar S. Use of 2D U-net convolutional neural networks for automated cartilage and meniscus segmentation of knee MR imaging data to determine relaxometry and morphometry. Radiology 2018;288(1):177–185. - PMC - PubMed

LinkOut - more resources