scholarly journals Detection and Classification of Knee Injuries from MR Images Using the MRNet Dataset with Progressively Operating Deep Learning Methods

2021 ◽  
Vol 3 (4) ◽  
pp. 1009-1029
Author(s):  
Ali Can Kara ◽  
Fırat Hardalaç
Keyword(s):  
Deep Learning ◽  
Cruciate Ligament ◽  
Knee Injuries ◽  
Mr Images ◽  
Learning Methods ◽  
Learning Group ◽  
Magnetic Resonance Imaging Mri ◽  
Anterior Cruciate ◽  
Novel Model

This study aimed to build progressively operating deep learning models that could detect meniscus injuries, anterior cruciate ligament (ACL) tears and knee abnormalities in magnetic resonance imaging (MRI). The Stanford Machine Learning Group MRNet dataset was employed in the study, which included MRI image indexes in the coronal, sagittal, and axial axes, each having 1130 trains and 120 validation items. The study is divided into three sections. In the first section, suitable images are selected to determine the disease in the image index based on the disturbance under examination. It is also used to identify images that have been misclassified or are noisy and/or damaged to the degree that they cannot be utilised for diagnosis in the first section. The study employed the 50-layer residual networks (ResNet50) model in this section. The second part of the study involves locating the region to be focused on based on the disturbance that is targeted to be diagnosed in the image under examination. A novel model was built by integrating the convolutional neural networks (CNN) and the denoising autoencoder models in the second section. The third section is dedicated to making a diagnosis of the disease. In this section, a novel ResNet50 model is trained to identify disease diagnoses or abnormalities, independent of the ResNet50 model used in the first section. The images that each model selects as output after training are referred to as progressively operating deep learning methods since they are supplied as an input to the following model.

scholarly journals Morphometric parameters as risk factors for anterior cruciate ligament injuries: A MRI case-control study

2014 ◽  
Vol 71 (3) ◽  
pp. 271-276 ◽  
Author(s):  
Lazar Stijak ◽  
Marko Bumbasirevic ◽  
Marko Kadija ◽  
Gordana Stankovic ◽  
Richard Herzog ◽  
...  
Keyword(s):  
Anterior Cruciate Ligament ◽  
Cruciate Ligament ◽  
Knee Injuries ◽  
Control Group ◽  
Study Group ◽  
Matched Pairs ◽  
Intercondylar Notch ◽  
Anterior Cruciate Ligament Injuries ◽  
Magnetic Resonance Imaging Mri ◽  
Anterior Cruciate

Background/Aim. The anterior cruciate ligament (ACL) is the most frequently injured ligament of the knee, representing 50% of all knee injuries. The aim of this study was to determine the differences in the morphometry of knee injury patients with an intact and a ruptured anterior cruciate ligament. Methods. The study included 33 matched pairs of patients divided into two groups: the study group with the diagnosis of anterior cruciate ligament rupture, and the control group with the diagnosis of patellofemoral pain but no anterior cruciate ligament lesion. The patients were matched on the basis of 4 attributes: age, sex, type of lesion (whether it was profession- related), and whether the lesion was left- or right-sided. Measurements were carried out using magnetic resonance imaging (MRI). Results. The anterior and posterior edges of the anterior cruciate ligament in the control group were highly significantly smaller (p < 0.01; in both cases). The control group showed a statistically significantly larger width of the anterior cruciate ligament (p < 0.05). A significant correlation between the width of the anterior cruciate ligament and the width (p < 0.01) and height (p < 0.05) of the intercondylar notch was found to exist in the control group, but not in the study group (p > 0.05). The patients in the control group showed a shorter but wider anterior cruciate ligament in comparison to their matched pairs. The control group of patients was also characterized by the correlation between the width of the intercondylar notch and the width of the anterior cruciate ligament, which was not the case in the study group. Conclusions. According to the results of our study we can say that a narrow intercondylar notch contains a proportionally thin anterior cruciate ligament, but we cannot say that this factor necessarily leads to rupture of the anterior cruciate ligament.

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

2019 ◽  
Vol 1 (3) ◽  
pp. 180091 ◽  
Author(s):  
Fang Liu ◽  
Bochen Guan ◽  
Zhaoye Zhou ◽  
Alexey Samsonov ◽  
Humberto Rosas ◽  
...  
Keyword(s):  
Anterior Cruciate Ligament ◽  
Deep Learning ◽  
Cruciate Ligament ◽  
Mr Images ◽  
Automated Diagnosis ◽  
Anterior Cruciate ◽  
Cruciate Ligament Tears

scholarly journals Correlation between Magnetic Resonance Imaging and Arthroscopic Findings in Knee Lesions

2021 ◽  
Vol 6 (3) ◽  
pp. 153-156
Author(s):  
Octav Russu ◽  
Andrei Feier ◽  
Emilian Ciorcila ◽  
Georgiana Miler ◽  
Cristian Trambitas ◽  
...  
Keyword(s):  
Magnetic Resonance Imaging ◽  
Magnetic Resonance ◽  
Cruciate Ligament ◽  
Correct Diagnosis ◽  
Knee Injuries ◽  
High Sensitivity ◽  
Kappa Coefficient ◽  
Resonance Imaging ◽  
Magnetic Resonance Imaging Mri ◽  
Anterior Cruciate

Abstract Background: Magnetic resonance imaging (MRI) is a noninvasive diagnostic method with no known side effects and a high sensitivity for detecting meniscal lesions as well as lesions of the anterior and posterior cruciate ligaments. For a correct and complete diagnosis, MRI results must be interpreted in the context of clinical examination and anamnesis. MRI results can support the surgical indication for arthroscopy, which is a minimally invasive intervention that facilitates the visualization, investigation, diagnosis, and treatment of intra-articular lesions of the knee. The purpose of this study was to assess the correlation between MRI results and the intra-articular lesions detected arthroscopically. Material and methods: This retrospective study evaluated a total of 60 patients admitted and treated between January 1, 2013 and January 1, 2014 in the Clinic of Orthopedics and Traumatology of Mureș County Hospital, Târgu Mureș, Romania. Results: In the 43 patients with lesion of the anterior cruciate ligament, the lesions were seen on MRI in 40 patients (93%) and confirmed arthroscopically in 37 patients (86%). In 34 cases (79.1% of the total of 43), the MRI and arthroscopic results were similar, the Kappa coefficient showing a high degree of correlation (0.90). Conclusion: By assessing the Kappa coefficient to highlight the correlation between MRI results and arthroscopic diagnosis for knee injuries, a statistically relevant correlation between the two methods was found. This suggests that a correct diagnosis can be achieved by both methods.

Classification of audio codecs with variable bit-rates using deep-learning methods

2020 ◽  
pp. 102952
Author(s):  
Atieh Khodadadi ◽  
Soheila Molaei ◽  
Mehdi Teimouri ◽  
Hadi Zare
Keyword(s):  
Deep Learning ◽  
Learning Methods

Deep Learning-Based Binary Classification of ADHD Using Resting State MR Images

2021 ◽  
Vol 6 (1) ◽  
Author(s):  
Vikas Khullar ◽  
Karuna Salgotra ◽  
Harjit Pal Singh ◽  
Davinder Pal Sharma
Keyword(s):  
Deep Learning ◽  
Resting State ◽  
Binary Classification ◽  
Mr Images

Diagnostic of Injury Risk in the Anterior Cruciate Ligament Based On Shape Context Description of the Intercondylar Notch Curvature

2021 ◽  
Author(s):  
Joao Paulo Dias ◽  
Ariful Bhuiyan ◽  
Nabila Shamim
Keyword(s):  
Anterior Cruciate Ligament ◽  
Injury Risk ◽  
Cruciate Ligament ◽  
Acl Injury ◽  
False Positives ◽  
Intercondylar Notch ◽  
Mr Images ◽  
Shape Context ◽  
Similarity Factor ◽  
Anterior Cruciate

Abstract An estimated number of 300,000 new anterior cruciate ligament (ACL) injuries occur each year in the United States. Although several magnetic resonance (MR) imaging-based ACL diagnostics methods have already been proposed in the literature, most of them are based on machine learning or deep learning strategies, which are computationally expensive. In this paper, we propose a diagnostics framework for the risk of injury in the anterior cruciate ligament (ACL) based on the application of the inner-distance shape context (IDSC) to describe the curvature of the intercondylar notch from MR images. First, the contours of the intercondylar notch curvature from 91 MR images of the distal end of the femur (70 healthy and 21 with confirmed ACL injury) were extracted manually using standard image processing tools. Next, the IDSC was applied to calculate the similarity factor between the extracted contours and reference standard curvatures. Finally, probability density functions of the similarity factor data were obtained through parametric statistical inference, and the accuracy of the ACL injury risk diagnostics framework was assessed using receiver operating characteristic analysis (ROC). The overall results for the area under the curve (AUC) showed that method reached a maximum accuracy of about 66%. Furthermore, the sensitivity and specificity results showed that an optimum discrimination threshold value for the similarity factor can be pursued that minimizes the incidence of false positives and false positives simultaneously.

scholarly journals Effects of a Prophylactic Knee Sleeve on Anterior Cruciate Ligament Loading During Sport-Specific Movements

2019 ◽  
Vol 28 (1) ◽  
pp. 1-7 ◽  
Author(s):  
Jonathan Sinclair ◽  
Paul J. Taylor
Keyword(s):  
Anterior Cruciate Ligament ◽  
Repeated Measures ◽  
Cruciate Ligament ◽  
Knee Injuries ◽  
Musculoskeletal Modeling ◽  
Acl Injuries ◽  
Recreational Athletes ◽  
Load Rate ◽  
Anterior Cruciate ◽  
3D Motion Capture

Context: Prophylactic knee bracing is extensively utilized in athletic populations to reduce the high risk from knee injuries, but its role in the attenuation of anterior cruciate ligament (ACL) pathologies is not well understood. Objective: The aim of this investigation was to explore the effects of a prophylactic knee sleeve on ACL loading parameters linked to the etiology of injury in recreational athletes. Setting: Laboratory. Design: Repeated measures. Participants: Thirteen healthy male recreational athletes. Intervention: Participants performed run, cut, and single-leg hop movements under 2 conditions; prophylactic knee sleeve and no sleeve. Main Outcome Measures: Biomechanical data were captured using an 8-camera 3D motion capture system and a force platform. Peak ACL force, average ACL load rate, and instantaneous ACL load rate were quantified using a musculoskeletal modeling approach. Results: The results showed that both average and instantaneous ACL load rates were significantly reduced when wearing the knee sleeve in the hop (sleeve = 612.45/1286.39 N/kg/s and no sleeve = 743.91/1471.42 N/kg/s) and cut (sleeve = 222.55/1058.02 N/kg/s and no sleeve = 377.38/1183.01 N/kg/s) movements. Conclusions: Given the biomechanical association between ACL loading and the etiology of ACL injuries, it is proposed that athletes may be able to attenuate their risk from injury during cut and hop movements through utilization of a prophylactic knee sleeve.

Sign in / Sign up

Export Citation Format

Share Document