Fully Automated Diagnosis of Anterior Cruciate Ligament Tears on Knee MR Images by Using Deep Learning
Fang Liu* , Bochen Guan*, Zhaoye Zhou, Alexey Samsonov, Humberto Rosas, Kevin Lian, Ruchi Sharma, Andrew Kanarek, John Kim, Ali Guermazi, Richard Kijowski
* F.L. and B.G. contributed equally to this work.

Author Affiliations
From the Department of Radiology, University of Wisconsin School of Medicine and Public Health, 600 Highland Avenue, Madison, WI 53705 (F.L., B.G., A.S., H.R., K.L., R.S., A.K., J.K., R.K.); Department of Electrical and Computer Engineering, University of Wisconsin School of Engineering, Madison, Wis (B.G.); Department of Biomedical Engineering, University of Minnesota, Minneapolis, Minn (Z.Z.); and Department of Radiology, Boston University School of Medicine, Boston, Mass (A.G.).
Address correspondence to F.L. (e-mail: fliu37@wisc.edu).
Published Online:May 8 2019https://doi.org/10.1148/ryai.2019180091

Abstract
There was no statistically significant difference between the anterior cruciate ligament (ACL) tear detection system and clinical radiologists with varying levels of experience for determining the presence or absence of a full-thickness ACL tear using sagittal proton density–weighted and fat-suppressed T2-weighted fast spin-echo MR images.

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, 2019