|
|
基于深度学习的无监督单模态医学图像配准算法
|
Abstract:
图像配准在手术导航、肿瘤监测等临床医学图像分析领域有着重要应用。本文针对现有无监督单模图像匹配算法的配准精度不够高的问题，提出了一种基于深度学习的无监督单模态医学图像配准算法。该方法引入短连接与长连接结合形成密集连接，改进了U-Net的特征图连接方式，解决相连接两特征图由于采样深度差距较大而产生较大语义差距的问题；在U型网络的解码器上设计部署了通道注意力机制，能够有效抑制噪音，产生更加光滑的形变场，从而进一步提高后续配准精度。在临床应用的单模态脑部核磁共振图像数据集上进行了训练及测试，结果表明，本文提出算法在配准精度上有了一定的提高。
Image registration has important applications in clinical medical image analysis fields such as sur-gical navigation and tumor monitoring. Aiming at the problem that the registration accuracy of the existing unsupervised single-modal image matching algorithm is not high enough, this paper proposes an unsupervised single-modal medical image registration algorithm based on deep learning. This method introduces the combination of short connections and long connections to form dense connections, improves the connection method of U-Net’s feature maps, and solves the problem of a large semantic gap between two connected feature maps due to the large difference in sampling depth; the channel attention mechanism is designed and deployed on the decoder of U-shaped network, which can effectively suppress noise and generate a smoother deformation field, thereby further improving the subsequent registration accuracy. The training and testing were carried out on the single-mode brain MRI image data set for clinical application. The results show that the algorithm proposed in this paper has a certain improvement in the registration accuracy.
| [1] | 谢姝婷. 基于深度学习的医学图像配准与分割研究[D]: [硕士学位论文]. 上海: 华东师范大学, 2022. |
| [2] | Fu, Y., Lei, Y., Wang, T., et al. (2020) Deep Learning in Medical Image Registration: A Review. Physics in Medicine & Biology, 65, 20TR01. https://doi.org/10.1088/1361-6560/ab843e |
| [3] | Rohé, M.-M., Datar, M., Heimann, T., et al. (2017) SVF-Net: Learning Deformable Image Registration Using Shape Matching. Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, Quebec City, 11-13 September 2017, 266-274. https://doi.org/10.1007/978-3-319-66182-7_31 |
| [4] | De Vos, B.D., Berendsen, F.F., Viergever, M.A., et al. (2019) A Deep Learning Framework for Unsupervised Affine and Deformable Image Registration. Medical Image Analysis, 52, 128-143.
https://doi.org/10.1016/j.media.2018.11.010 |
| [5] | Li, H. and Fan, Y. (2018) Non-Rigid Image Registration Using Self-Supervised Fully Convolutional Networks without Training Data. Proceedings of the 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018), Washington DC, 4-7 April 2018, 1075-1078. https://doi.org/10.1109/ISBI.2018.8363757 |
| [6] | Yang, X., Kwitt, R. and Niethammer, M. (2016) Fast Predictive Image Registration. In: Deep Learning and Data Labeling for Medical Applications, Springer, Berlin, 48-57. https://doi.org/10.1007/978-3-319-46976-8_6 |
| [7] | Sokooti, H., Vos, B.D., Berendsen, F., et al. (2017) Nonrigid Image Registration Using Multi-Scale 3D Convolutional Neural Networks. Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, Quebec City, 11-13 September 2017, 232-239. https://doi.org/10.1007/978-3-319-66182-7_27 |
| [8] | Balakrishnan, G., Zhao, A., Sabuncu, M.R., et al. (2019) VoxelMorph: A Learning Framework for Deformable Medical Image Registration. IEEE Transactions on Medical Im-aging, 38, 1788-1800.
https://doi.org/10.1109/TMI.2019.2897538 |
| [9] | Ronneberger, O., Fischer, P. and Brox, T. (2015) U-Net: Convo-lutional Networks for Biomedical Image Segmentation. Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention, Munich, 5-9 October 2015, 234-241. https://doi.org/10.1007/978-3-319-24574-4_28 |
| [10] | Jaderberg, M., Simonyan, K. and Zisserman, A. (2015) Spatial Transformer Networks. Proceedings of the 28th International Conference on Neural Information Processing Systems, Volume 2, 2017-2025. |
| [11] | Tian, F., Gao, Y., Fang, Z., et al. (2021) Depth Estimation Using a Self-Supervised Net-work Based on Cross-Layer Feature Fusion and the Quadtree Constraint. IEEE Transactions on Circuits and Systems for Video Technology, 32, 1751-1766. https://doi.org/10.1109/TCSVT.2021.3080928 |
| [12] | Gao, Y., Liu, X., Li, J., et al. (2022) LFT-Net: Local Feature Transformer Network for Point Clouds Analysis. IEEE Transactions on Intelligent Transportation Systems. https://doi.org/10.1109/TITS.2022.3140355 |
| [13] | Oh, D., Kim, B., Lee, J., et al. (2021) Unsupervised Deep Learning Network with Self-Attention Mechanism for Non-Rigid Registration of 3D Brain MR Im-ages. Journal of Medical Imaging and Health Informatics, 11, 736-751.
https://doi.org/10.1166/jmihi.2021.3345 |
| [14] | Wang, Q., Wu, B., Zhu, P., et al. (2019) ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks. Computer Vision and Pattern Recognition, Long Beach, 15-20 June 2019, 11532-11542.
https://doi.org/10.1109/CVPR42600.2020.01155 |
| [15] | Marcus, D.S., Wang, T.H., Parker, J., et al. (2007) Open Access Series of Imaging Studies (OASIS): Cross-Sectional MRI Data in Young, Middle Aged, Nondemented, and Demented Older Adults. Journal of Cognitive Neuroscience, 19, 1498-1507. https://doi.org/10.1162/jocn.2007.19.9.1498 |
| [16] | Fischl, B. (2012) FreeSurfer. Neuroimage, 62, 774-781. https://doi.org/10.1016/j.neuroimage.2012.01.021 |
| [17] | Collins, D.L., Zijdenbos, A.P., Kollokian, V., et al. (1998) Design and Construction of a Realistic Digital Brain Phantom. IEEE Transactions on Medical Imaging, 17, 463-468. https://doi.org/10.1109/42.712135 |
| [18] | Avants, B.B., Tustison, N.J., Song, G., et al. (2011) A Reproducible Eval-uation of ANTs Similarity Metric Performance in Brain Image Registration. Neuroimage, 54, 2033-2044. https://doi.org/10.1016/j.neuroimage.2010.09.025 |
