Department of Biomedical Engineering, University of Minnesota, Minneapolis, Minnesota.
Departments of Radiology, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin.
Magn Reson Med. 2018 Dec;80(6):2759-2770. doi: 10.1002/mrm.27229. Epub 2018 May 17.
To describe and evaluate a new segmentation method using deep convolutional neural network (CNN), 3D fully connected conditional random field (CRF), and 3D simplex deformable modeling to improve the efficiency and accuracy of knee joint tissue segmentation.
A segmentation pipeline was built by combining a semantic segmentation CNN, 3D fully connected CRF, and 3D simplex deformable modeling. A convolutional encoder-decoder network was designed as the core of the segmentation method to perform high resolution pixel-wise multi-class tissue classification for 12 different joint structures. The 3D fully connected CRF was applied to regularize contextual relationship among voxels within the same tissue class and between different classes. The 3D simplex deformable modeling refined the output from 3D CRF to preserve the overall shape and maintain a desirable smooth surface for joint structures. The method was evaluated on 3D fast spin-echo (3D-FSE) MR image data sets. Quantitative morphological metrics were used to evaluate the accuracy and robustness of the method in comparison to the ground truth data.
The proposed segmentation method provided good performance for segmenting all knee joint structures. There were 4 tissue types with high mean Dice coefficient above 0.9 including the femur, tibia, muscle, and other non-specified tissues. There were 7 tissue types with mean Dice coefficient between 0.8 and 0.9 including the femoral cartilage, tibial cartilage, patella, patellar cartilage, meniscus, quadriceps and patellar tendon, and infrapatellar fat pad. There was 1 tissue type with mean Dice coefficient between 0.7 and 0.8 for joint effusion and Baker's cyst. Most musculoskeletal tissues had a mean value of average symmetric surface distance below 1 mm.
The combined CNN, 3D fully connected CRF, and 3D deformable modeling approach was well-suited for performing rapid and accurate comprehensive tissue segmentation of the knee joint. The deep learning-based segmentation method has promising potential applications in musculoskeletal imaging.
描述并评估一种新的分割方法，该方法使用深度卷积神经网络（CNN）、3D 全连接条件随机场（CRF）和 3D 单形变形建模来提高膝关节组织分割的效率和准确性。
通过结合语义分割 CNN、3D 全连接 CRF 和 3D 单形变形建模，构建了一个分割流水线。设计了一个卷积编码器-解码器网络作为分割方法的核心，对 12 种不同的关节结构进行高分辨率像素级多类组织分类。3D 全连接 CRF 用于正则化同一组织类别内和不同类别之间体素之间的上下文关系。3D 单形变形建模细化了 3D CRF 的输出，以保持关节结构的整体形状并保持理想的光滑表面。该方法在 3D 快速自旋回波（3D-FSE）MR 图像数据集上进行了评估。使用定量形态学指标来评估该方法与地面真实数据相比的准确性和稳健性。
所提出的分割方法在分割所有膝关节结构方面表现良好。有 4 种组织类型的平均 Dice 系数高于 0.9，包括股骨、胫骨、肌肉和其他未指定的组织。有 7 种组织类型的平均 Dice 系数在 0.8 到 0.9 之间，包括股骨软骨、胫骨软骨、髌骨、髌骨软骨、半月板、股四头肌和髌腱以及髌下脂肪垫。1 种组织类型的关节积液和贝克囊肿的平均 Dice 系数在 0.7 到 0.8 之间。大多数肌肉骨骼组织的平均对称面距离值低于 1mm。
联合使用 CNN、3D 全连接 CRF 和 3D 变形建模方法非常适合快速准确地对膝关节进行全面的组织分割。基于深度学习的分割方法在肌肉骨骼成像中有很好的应用潜力。
