3D convolutional neural networks for detection and severity staging of meniscus and PFJ cartilage morphological degenerative changes in osteoarthritis and anterior cruciate ligament subjects.
3D convolutional neural networks for detection and severity staging of meniscus and PFJ cartilage morphological degenerative changes in osteoarthritis and anterior cruciate ligament subjects.
复制标题
DOI：
10.1002/jmri.26246
复制
发表时间：
2019-03
期刊：
影响因子：
--
通讯作者：
Majumdar S
中科院分区：
文献类型：
--
作者：
Pedoia V;Norman B;Mehany SN;Bucknor MD;Link TM;Majumdar S
MRI的半定量评估在肌肉骨骼研究中起着核心作用;然而，在临床环境中，MRI报告往往是主观和定性的。在研究中使用的分级方案没有使用，因为它们非常耗时，在临床实践中不可行。评估深度学习模型检测骨关节炎和前交叉韧带（ACL）受试者半月板和髌股软骨病变严重程度并对其进行分期的能力。回顾性研究旨在评价技术发展。总共1478项MRI研究，包括骨关节炎不同阶段和ACL损伤和重建后的受试者。3T MRI，3D FSE CUBE。使用2D U-Net自动分割软骨和半月板，使用3D卷积神经网络（3D-CNN）自动检测半月板和软骨病变的严重程度。受试者工作特征（ROC）曲线，特异性和敏感性，以及分类准确性。半月板病变的敏感性和特异性分别为89.81%和81.98%，软骨病变的敏感性和特异性分别为80.0%和80.27%。通过纳入人口统计学因素，获得了病变严重程度分期的最佳性能，正常、小和复杂大病变的准确率分别为80.74%、78.02%和75.00%。在这项研究中，我们提供了一个完全自动化的深度学习管道的概念证明，可以识别髌骨和髌骨软骨病变的存在。该管道还显示出对病变受试者进行更深入检查以进行多分类预测和严重程度分期的潜力。2阶段2
Semiquantitative assessment of MRI plays a central role in musculoskeletal research; however, in the clinical setting MRI reports often tend to be subjective and qualitative. Grading schemes utilized in research are not used because they are extraordinarily time-consuming and unfeasible in clinical practice. To evaluate the ability of deep-learning models to detect and stage severity of meniscus and patellofemoral cartilage lesions in osteoarthritis and anterior cruciate ligament (ACL) subjects. Retrospective study aimed to evaluate a technical development. In all, 1478 MRI studies, including subjects at various stages of osteoarthritis and after ACL injury and reconstruction. 3T MRI, 3D FSE CUBE. Automatic segmentation of cartilage and meniscus using 2D U-Net, automatic detection, and severity staging of meniscus and cartilage lesion with a 3D convolutional neural network (3D-CNN). Receiver operating characteristic (ROC) curve, specificity and sensitivity, and class accuracy. Sensitivity of 89.81% and specificity of 81.98% for meniscus lesion detection and sensitivity of 80.0% and specificity of 80.27% for cartilage were achieved. The best performances for staging lesion severity were obtained by including demographics factors, achieving accuracies of 80.74%, 78.02%, and 75.00% for normal, small, and complex large lesions, respectively. In this study we provide a proof of concept of a fully automated deep-learning pipeline that can identify the presence of meniscal and patellar cartilage lesions. This pipeline has also shown potential in making more in-depth examinations of lesion subjects for multiclass prediction and severity staging. 2 Stage 2
登录
查看更多内容
影响因子：
4.4
作者：
Lee H;Tajmir S;Lee J;Zissen M;Yeshiwas BA;Alkasab TK;Choy G;Do S
通讯作者：
Do S
一种基于回归系数的高效弹性网络光谱数据变量选择方法
影响因子：
3.7
作者：
Liu W;Li Q
通讯作者：
Li Q
影响因子：
3.3
作者：
Chaudhari AS;Fang Z;Kogan F;Wood J;Stevens KJ;Gibbons EK;Lee JH;Gold GE;Hargreaves BA
通讯作者：
Hargreaves BA
影响因子：
22.7
作者：
Lee, Honglak;Grosse, Roger;Ng, Andrew Y.
通讯作者：
Ng, Andrew Y.
影响因子：
2.1
作者：
Joseph, Gabby B.;Hou, Stephanie W.;Link, Thomas M.
通讯作者：
Link, Thomas M.