基于深度神经网络的尺桡骨远端图像语义分割

DRU Image Semantic Segmentation Using Deep Neural Networks

摘要: 对尺桡骨远端图像进行语义分割可以提取出尺骨桡骨感兴趣区域(ROI)，ROI提取对后期的辅助诊断分析非常重要。本文提出了一种基于深度全卷积神经网络的语义分割模型，并对尺桡骨远端图像进行像素级的语义分割。实验表明，该模型可以精准地从尺桡骨远端图像中分割尺骨、桡骨并识别其语义，并分析了几种典型网络结构对分割模型的影响作用。分割模型对尺骨的识别精度和召回率为97%，交并比(IoU)为95%；对桡骨识别精度为98.5%，召回率为98%，交并比(IoU)为96.6%。

Abstract: Abstract Semantic segmentation of the distal radius and ulna images can extract the region of interest (ROI) for Ulnar radius, which counts a great deal for bone age recognition. In this work, we propose a fully convolutional network based model for semantic segmentation. Experimental results show that the proposed model can precisely segment the ulna and radius from distal radius and ulna images. We also discussed the influence of different network structures. For ulna, we get an accuracy of 97%, recall of 97%, IoU of 95%. For radius, we get an accuracy of 98.5%, recall of 98%, IoU of 96.6%.

文章引用：胡明辉, 李俊, 申妍燕, 曹松, 王书强. 基于深度神经网络的尺桡骨远端图像语义分割[J]. 图像与信号处理, 2018, 7(2): 85-95. https://doi.org/10.12677/JISP.2018.72010

参考文献

[1]	张新峰, 沈兰荪. 图像分割技术研究[J]. 电路与系统学报, 2004, 9(2):92-99.
[2]	杨晖, 曲秀杰. 图像分割方法综述[J]. 电脑开发与应用, 2005, 18(3):21-23.
[3]	Felzenszwalb, P.F. and Huttenlocher, D.P. (2004) Efficient Graph-Based Image Segmentation. International Journal of Computer Vision, 59, 167-181. [Google Scholar] [CrossRef]
[4]	Lecun, Y., Bottou, L., Bengio, Y., et al. (1998) Gradient-Based Learning Applied to Document Recognition. Proceedings of the IEEE, 86, 2278-2324. [Google Scholar] [CrossRef]
[5]	Oe, S. and Liu, G. (2000) Texture Image Segmentation Method Based on Multilayer CNN. International Conference on Tools with Artificial Intelligence, 31, 147-150.
[6]	Shimoda, W. and Yanai, K. (2015) CNN-Based Food Image Segmentation without Pixel-Wise Annotation. International Conference on Image Analysis and Processing, 9281, 449-457.
[7]	Vetrivel, A., Kerle, N., Gerke, M., et al. (2016) Towards Automated Satellite Image Segmentation and Classi-fication for Assessing Disaster Damage Using Data-Specific Features with Incremental Learning. GEOBIA 2016: Solutions and Syn-ergies, 14-16 September 2016. [Google Scholar] [CrossRef]
[8]	Rahmat, B., Joelianto, E., Purnama, I.K.E., et al. (2013) CNN-Fuzzy-Based Indonesia License Plate Image Segmentation System. IEEE International Conference on Computer Science and Automation Engineering, Guangzhou.
[9]	Greenspan, H., Ginneken, B.V. and Summers, R.M. (2016) Guest Editorial Deep Learning in Medical Imaging: Overview and Future Promise of an Exciting New Technique. IEEE Transactions on Medical Imaging, 35, 1153-1159. [Google Scholar] [CrossRef]
[10]	Rouhi, R., Jafari, M., Kasaei, S., et al. (2015) Benign and Malignant Breast Tumors Classification Based on Region Growing and CNN Segmentation. Expert Systems with Applications, 42, 990-1002. [Google Scholar] [CrossRef]
[11]	Perfetti, R., Ricci, E., Casali, D., et al. (2008) A CNN Based Algorithm for Retinal Vessel Segmentation. 12th WSEAS International Conference on CIRCUITS, Heraklion, 22-24 July 2008, 152-157.
[12]	Duraisamy, M. and Duraisamy, S. (2012) CNN-Based Approach for Segmentation of Brain and Lung MRI Images. European Journal of Scientific Research, 81, 298-313.
[13]	Moeskops, P., Viergever, M.A., Mendrik, A.M., et al. (2016) Automatic Segmentation of MR Brain Images with a Convolutional Neural Network. IEEE Transactions on Medical Imaging, 35, 1252-1261. [Google Scholar] [CrossRef]
[14]	Lotter, W., Sorensen, G. and Cox, D. (2017) A Multi-Scale CNN and Curri-culum Learning Strategy for Mammogram Classification. In: Cardoso, M., et al., Eds., Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support. DLMIA 2017, ML-CDS 2017. Lecture Notes in Computer Science, Vol. 10553. Springer, Cham.
[15]	Chen, L.C., Papandreou, G., Kokkinos, I., et al. (2014) Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs. Computer Science, 4, 357-361.
[16]	Long, J., Shelhamer, E. and Darrell, T. (2015) Fully Convolutional Networks for Semantic Segmentation. IEEE Conference on Computer Vision and Pattern Recognition, IEEE Computer Society, 3431-3440.
[17]	Ronneberger, O., Fischer, P. and Brox, T. (2015) U-Net: Convolutional Networks for Biomedical Image Segmentation. International Conference on Medical Image Computing and Computer-Assisted Intervention, Springer, Cham, 234-241.
[18]	Yang, L., Rodriguez, H., Crucianu, M. and Ferecatu, M. (2017) Fully Convolutional Network with Superpixel Parsing for Fashion Web Image Segmentation. In: Amsaleg, L., Guðmundsson, G., Gurrin, C., Jónsson, B. and Satoh, S., Eds., MultiMedia Modeling. MMM 2017. Lecture Notes in Computer Science, Vol. 10132. Springer, Cham.
[19]	Niemeijer, J., Fouopi, P.P., Knake-Langhorst, S., et al. (2017) A Review of Neural Network based Semantic Segmentation for Scene Understanding in Context of the self driving Car. BioMedTec Studierendentagung.
[20]	Luk, K.D., Saw, L.B., Grozman, S., et al. (2014) Assessment of Skeletal Maturity in Scoliosis Patients to Determine Clinical Management: A New Classification Scheme Using Distal Radius and Ulna Radiographs. Spine Journal Official Journal of the North American Spine Society, 14, 315-325. [Google Scholar] [CrossRef] [PubMed]
[21]	Simonyan, K. and Zisserman, A. (2015) Very Deep Convolutional Networks for Large-Scale Image Recognition. In: International Conference on Learning Representations (ICLR).
[22]	Torralba, A., Russell, B.C. and Yuen, J. (2010) LabelMe: Online Image Annotation and Applications. Proceedings of the IEEE, 98, 1467-1484. [Google Scholar] [CrossRef]
[23]	李航. 统计学习方法[M]. 北京: 清华大学出版社, 2012.
[24]	Deng, J., Dong, W., Socher, R., et al. (2009) ImageNet: A Large-Scale Hierarchical Image Database. IEEE Conference on Computer Vision and Pattern Recognition, Miami, 20-25 June 2009, 248-255.