Automatic detection and voxel-wise mapping of lumbar spine Modic changes with deep learning.
Automatic detection and voxel-wise mapping of lumbar spine Modic changes with deep learning.
复制标题
DOI：
10.1002/jsp2.1204
复制
发表时间：
2022-06
期刊：
影响因子：
3.7
通讯作者：
Majumdar, Sharmila
中科院分区：
文献类型：
--
作者：
Gao, Kenneth T.;Tibrewala, Radhika;Hess, Madeline;Bharadwaj, Upasana U.;Inamdar, Gaurav;Link, Thomas M.;Chin, Cynthia T.;Pedoia, Valentina;Majumdar, Sharmila
Modic 变化 (MC) 是描述椎骨磁共振成像 (MRI) 信号强度变化的最流行的分类系统。然而，由于缺乏确凿的证据证明这些异常与腰痛的关联，因此越来越需要新的定量和标准化方法来表征这些异常，特别是对于过渡性或混合性病变。这项回顾性成像研究旨在开发一种可解释的基于深度学习的检测工具，用于 MC 的体素映射。 75 例腰椎 MRI 检查显示急性至慢性腰痛、神经根病和其他腰椎症状。该管道由两个深度卷积神经网络组成，用于生成可解释的体素方式 Modic 图。首先，训练自动编码器从 T1 加权矢状腰椎图像中分割椎体。接下来，两名放射科医生从 T1 和 T2 加权组合评估中对 MC 进行分割和标记，作为训练第二个自动编码器执行 MC 分割的基本事实。然后使用基于规则的信号强度算法将检测到的区域中的体素分类为适当的 Modic 类型。事后，三名放射科医生在人工 (AI) 辅助实验中借助模型预测独立对第二个数据集进行了评分。该模型成功识别出未见过的测试集中 85.7% 的样本存在变化，灵敏度为 0.71 (±0.072)，特异性为 0.95 (±0.022)，Cohen 的 kappa 得分为 0.63。在人工智能辅助实验中，初级放射科医生和高级神经放射科医生之间的一致性从 Cohen 的 kappa 评分 0.52 显着提高到 0.58（p< 0.05）。这种基于深度学习的方法与放射科医生达成了实质性共识，并且可以作为提高 MC 评估中评估者间可靠性的工具。 Modic 变化是描述椎骨 MRI 信号强度变化的最流行的分类系统。这项回顾性成像研究开发了一种可解释的基于深度学习的检测工具，用于 Modic 变化的体素映射，该工具与放射科医生的观点基本一致。
Modic changes (MCs) are the most prevalent classification system for describing magnetic resonance imaging (MRI) signal intensity changes in the vertebrae. However, there is a growing need for novel quantitative and standardized methods of characterizing these anomalies, particularly for lesions of transitional or mixed nature, due to the lack of conclusive evidence of their associations with low back pain. This retrospective imaging study aims to develop an interpretable deep learning‐based detection tool for voxel‐wise mapping of MCs. Seventy‐five lumbar spine MRI exams that presented with acute‐to‐chronic low back pain, radiculopathy, and other symptoms of the lumbar spine were enrolled. The pipeline consists of two deep convolutional neural networks to generate an interpretable voxel‐wise Modic map. First, an autoencoder was trained to segment vertebral bodies from T1‐weighted sagittal lumbar spine images. Next, two radiologists segmented and labeled MCs from a combined T1‐ and T2‐weighted assessment to serve as ground truth for training a second autoencoder that performs segmentation of MCs. The voxels in the detected regions were then categorized to the appropriate Modic type using a rule‐based signal intensity algorithm. Post hoc, three radiologists independently graded a second dataset with the aid of the model predictions in an artificial (AI)‐assisted experiment. The model successfully identified the presence of changes in 85.7% of samples in the unseen test set with a sensitivity of 0.71 (±0.072), specificity of 0.95 (±0.022), and Cohen's kappa score of 0.63. In the AI‐assisted experiment, the agreement between the junior radiologist and the senior neuroradiologist significantly improved from Cohen's kappa score of 0.52 to 0.58 (p < 0.05). This deep learning‐based approach demonstrates substantial agreement with radiologists and may serve as a tool to improve inter‐rater reliability in the assessment of MCs. Modic changes are the most prevalent classification system for describing MRI signal intensity changes in the vertebrae. This retrospective imaging study develops an interpretable deep learning‐based detection tool for voxel‐wise mapping of Modic changes that demonstrates substantial agreement with radiologists.
登录
查看更多内容
与腰痛和活动限制的敏捷性变化相关：系统文献综述和荟萃分析。
影响因子：
3.7
作者：
Herlin C;Kjaer P;Espeland A;Skouen JS;Leboeuf-Yde C;Karppinen J;Niinimäki J;Sørensen JS;Storheim K;Jensen TS
通讯作者：
Jensen TS
影响因子：
5.9
作者：
Farshad-Amacker, Nadja A.;Hughes, Alexander;Farshad, Mazda
通讯作者：
Farshad, Mazda
影响因子：
19.7
作者：
Hallinan, James Thomas Patrick Decourcy;Zhu, Lei;Quek, Swee Tian
通讯作者：
Quek, Swee Tian
影响因子：
3
作者：
Hutton, Michael J.;Bayer, Jens H.;Sharp, David J.
通讯作者：
Sharp, David J.
影响因子：
3.3
作者：
Karampinos, Dimitrios C.;Melkus, Gerd;Baum, Thomas;Bauer, Jan S.;Rummeny, Ernst J.;Krug, Roland
通讯作者：
Krug, Roland