Dataset Information

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

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.

SUBMITTER: Liu F

PROVIDER: S-EPMC6542618 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Purpose: To develop and evaluate a deep learning-based method to localize and classify anterior cruciate ligament (ACL) ruptures on knee MR images by using arthroscopy as the reference standard. Methods: We proposed a fully automated ACL rupture localization system to localize and classify ACL ruptures. The classification of ACL ruptures was based on the projection coordinates of the ACL rupture point on the line connecting the center coordinates of the femoral and tibial footprints. The line was divided into three equal parts and the position of the projection coordinates indicated the classification of the ACL ruptures (femoral side, middle and tibial side). In total, 85 patients (mean age: 27; male: 56) who underwent ACL reconstruction surgery under arthroscopy were included. Three clinical readers evaluated the datasets separately and their diagnostic performances were compared with those of the model. The performance metrics included the accuracy, error rate, sensitivity, specificity, precision, and F1-score. A one-way ANOVA was used to evaluate the performance of the convolutional neural networks (CNNs) and clinical readers. Intraclass correlation coefficients (ICC) were used to assess interobserver agreement between the clinical readers. Results: The accuracy of ACL localization was 3.77 ± 2.74 and 4.68 ± 3.92 (mm) for three-dimensional (3D) and two-dimensional (2D) CNNs, respectively. There was no significant difference in the ACL rupture location performance between the 3D and 2D CNNs or among the clinical readers (Accuracy, p < 0.01). The 3D CNNs performed best among the five evaluators in classifying the femoral side (sensitivity of 0.86 and specificity of 0.79), middle side (sensitivity of 0.71 and specificity of 0.84) and tibial side ACL rupture (sensitivity of 0.71 and specificity of 0.99), and the overall accuracy for sides classifying of ACL rupture achieved 0.79. Conclusion: The proposed deep learning-based model achieved high diagnostic performances in locating and classifying ACL fractures on knee MR images.

Project description:BackgroundAn anterior cruciate ligament (ACL) tear is a common knee injury, particularly among young and active individuals. Little is known, however, about the societal impacts of ACL tears, which could be large given the typical patient age and increased lifetime risk of knee osteoarthritis. This study evaluates the cost-effectiveness of ACL reconstruction compared with structured rehabilitation only.MethodsA cost-utility analysis of ACL reconstruction compared with structured rehabilitation only was conducted with use of a Markov decision model over two time horizons: the short to intermediate term (six years), on the basis of Level-I evidence derived from the KANON Study and the Multicenter Orthopaedic Outcomes Network (MOON) database; and the lifetime, on the basis of a comprehensive literature review. Utilities were assessed with use of the SF-6D. Costs (in 2012 U.S. dollars) were estimated from the societal perspective and included the effects of the ACL tear on work status, earnings, and disability. Effectiveness was expressed as quality-adjusted life years (QALYs) gained.ResultsIn the short to intermediate term, ACL reconstruction was both less costly (a cost reduction of $4503) and more effective (a QALY gain of 0.18) compared with rehabilitation. In the long term, the mean lifetime cost to society for a typical patient undergoing ACL reconstruction was $38,121 compared with $88,538 for rehabilitation. ACL reconstruction resulted in a mean incremental cost savings of $50,417 while providing an incremental QALY gain of 0.72 compared with rehabilitation. Effectiveness gains were driven by the higher probability of an unstable knee and associated lower utility in the rehabilitation group. Results were most sensitive to the rate of knee instability after initial rehabilitation.ConclusionsACL reconstruction is the preferred cost-effective treatment strategy for ACL tears and yields reduced societal costs relative to rehabilitation once indirect cost factors, such as work status and earnings, are considered. The cost of an ACL tear over the lifetime of a patient is substantial, and resources should be directed to developing innovations for injury prevention and for altering the natural history of an ACL injury.

Project description:PurposeTo compare hamstring autograft primary anterior cruciate ligament reconstruction (ACLR) techniques including adult-type/anatomic, transphyseal, and transphyseal techniques by (1) ACL graft tear, (2) contralateral ACL tear, and (3) all-cause ipsilateral reoperation.MethodsA retrospective, single-surgeon review was performed including all ACLR with hamstring autograft in pediatric and adolescent patients from 2011 to 2019. Minimum 2-year follow-up was required for patients unless a tear or reoperation was sustained before that time point. Data collected included demographics and baseline surgical variables, type of reconstruction, sporting activity, and deviations from rehabilitation protocols. Comparisons were made among hamstring autograft reconstruction groups (adult-type/anatomic, transphyseal, and partial transphyseal) for primary outcomes of graft tear, contralateral ACL tears, and all-cause ipsilateral knee reoperations, including hardware removal. Secondary surgeries performed with different surgeons were noted.ResultsIn total, 214 patients of age 15.2 ± 2.0 years with 4.1 ± 1.7-year follow-up were included. Overall graft tear rate was 11.7% (11.0% adult-type vs 19.1% transphyseal vs 5.6% partial transphyseal; P = .18). On univariate analyses, all-cause ipsilateral reoperation did not differ by technique (21.3% vs 31.0% vs 33.3%; P = .20), and neither did contralateral ACL tear (8.1% vs 9.5% vs 0%; P = .17). 21.7% of ipsilateral revision ACLRs (all adult-type) and 16.7% of patients with any reoperations had subsequent procedures performed with a different surgeon.ConclusionsThe graft tear rates in primary hamstring autograft ACLRs in the adolescent population did not significantly differ by technique (11.0% vs 19.1% vs 5.6% in adult-type, transphyseal, and partial transphyseal reconstructions, respectively). Furthermore, contralateral ACL tears (8.1% vs 9.5% vs 0%) and all-cause (including > 1/4 hardware removal) ipsilateral knee reoperations (21.3% vs 31.0% vs 33.3%) did not statistically differ. Higher powered studies may detect statistical significance in the observed differences in this study.Level of evidenceLevel IV, therapeutic case series.

Project description:BackgroundWe aimed to evaluate clinical and radiological results after simultaneous open-wedge high tibial osteotomy (HTO) and anterior cruciate ligament (ACL) reconstruction in patients with ACL deficiency combined with medial uni-compartmental osteoarthritis (OA) and varus deformity.MethodsThis retrospective study was performed using data collected from 2005 to 2011 on a total of 24 patients who were diagnosed with ACL injury and medial unicompartmental OA with varus deformity, and who subsequently underwent simultaneous open-wedge HTO and arthroscopic ACL reconstruction. The mean follow-up duration was 5.2 years. For clinical outcomes, we evaluated Lysholm score, Tegner activity score, range of motion, Lachmann test, and pivot-shift test, and for radiological outcomes, we evaluated the degree of varus deformity, progression of medial OA, tibial posterior slope, anterior instability, and postoperative complication.ResultsThere were no limitations in range of motion found in any cases. Three patients showed progressive osteoarthritis on the medial compartment. The mechanical femorotibial angle was significantly corrected from varus 7.0 degrees to valgus 1.2 degrees, and the tibial posterior slope was not significantly changed. The Lysholm and Tegner activity scores were significantly improved after surgery (from 58 to 94 points on the Lysholm scale and from 4.0 to 5.3 points on the Tegner activity scale). Although the Lachman test and the pivot-shift test showed significant improvements after surgery, instability greater than Gr II was observed in three patients on the Lachman test and in four patients on the pivot-shift test. The side-to-side difference improved from 9.6 mm to 4.2 mm postoperatively as assessed using a Telos® arthrometer. There were no cases of nonunion or fixation loss.ConclusionsSimultaneous open-wedge HTO and ACL reconstruction in patients with ACL injury with medial compartmental OA showed satisfactory functional outcomes and postoperative activity level scores. However, some patients showed residual instability and progression of OA.

Project description:Anterior cruciate ligament (ACL) deficiency alters 6 degrees of freedom knee kinematics, yet only anterior translation and internal rotation have been the primary measures in previous studies.To compare the 6 degrees of freedom knee kinematics and the graft forces after single- and double-bundle ACL reconstructions under various external loading conditions.Controlled laboratory study.Ten human cadaveric knees were tested with a robotic testing system under 4 conditions: intact, ACL deficient, single-bundle reconstructed with a quadrupled hamstring tendon graft, and double-bundle reconstructed with 2 looped hamstring tendon grafts. Knee kinematics and forces of the ACL or ACL graft in each knee were measured under 3 loading conditions: an anterior tibial load of 134 N, a simulated quadriceps muscle load of 400 N, and combined tibial torques (10 N.m valgus and 5 N.m internal tibial torques) at 0 degrees , 15 degrees , 30 degrees , 60 degrees , and 90 degrees of knee flexion.The double-bundle reconstruction restored the anterior and medial laxities closer to the intact knee than the single-bundle reconstruction. However, the internal rotation of the tibia under the simulated quadriceps muscle load was significantly decreased when compared with the intact knee after both reconstructions, more so after double-bundle reconstruction (P < .05). The entire graft force of the double-bundle reconstruction was more similar to that of the intact ACL than that of the single-bundle reconstruction. However, the posterolateral bundle graft in the double-bundle reconstructed knee was overloaded as compared with the intact posterolateral bundle.The double-bundle reconstruction can better restore the normal anterior-posterior and medial-lateral laxities than the single-bundle reconstruction can, but an overloading of the posterolateral bundle graft can occur in a double-bundle reconstructed knee.Both single-bundle and double-bundle techniques cannot restore the rotational laxities and the ACL force distributions of the intact knee.

Dataset Information

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

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets