Academic Journal of Computing & Information Science, 2023, 6(9); doi: 10.25236/AJCIS.2023.060915.
Haoxuan Sun1,2, Xiaogang Du1,2
1Shaanxi Joint Laboratory of Artificial Intelligence, Shaanxi University of Science and Technology, Xi’an, 710021, China
2The School of Electronic Information and Artificial Intelligence, Shaanxi University of Science and Technology, Xi’an, 710021, China
Aiming at the problem of the loss of detail information in medical image registration, which leads to poor registration results in large deformation regions, a large deformation features guided network(LDGNet) for medical image registration is proposed. LDGNet mainly includes two contributions: first, a large deformation feature enhancement module is designed at the encoding and decoding connection to enable the network to enhance the extraction of large deformation features. Secondly, a large deformation feature guidance module is designed at the skip connection, which can help fully fuse the large deformation features from the encoded feature map, and effectively improve the registration accuracy of the network in large deformation regions. Registration experiments on the brain dataset IXI show that LDGNet achieves higher registration accuracy compared with current popular medical image registration methods.
deep learning, deformable registration, unsupervised learning, large deformation feature enhancement, large deformation feature guidance
Haoxuan Sun, Xiaogang Du. Large Deformation Features Guided Network for Medical Image Registration. Academic Journal of Computing & Information Science (2023), Vol. 6, Issue 9: 102-109. https://doi.org/10.25236/AJCIS.2023.060915.
[1] Haskins G, Kruger U, Yan P. Deep learning in medical image registration: a survey[J]. Machine Vision and Applications, 2020, 31: 1-18.
[2] Sun W, Niessen W J, Klein S. Randomly perturbed B-splines for nonrigid image registration[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(7): 1401-1413.
[3] Sharifi H, Zhang H, Bagher-Ebadian H, et al. Utilization of a hybrid finite-element based registration method to quantify heterogeneous tumor response for adaptive treatment for lung cancer patients[J]. Physics in Medicine and Biology, 2018, 63(6): 065017.
[4] Bob D, Berendsen F F, Viergever M A. A deep learning framework for unsupervised affine and deformable image registration[J]. Medical Image Analysis, 2019, 52: 128-143.
[5] Cao X H, Yang J H, Zhang J, et al. Deformable image registration using a cue-aware deep regression network[J]. IEEE Transactions on Biomedical Engineering, 2018, 4(9): 1900-1911.
[6] Sokooti H, Vos B D, Berendsen F, et al. Nonrigid image registration using multi-scale 3D convolutional neural networks[C]//Proceeding of International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI). 2017: 232-239.
[7] Krebs J, Delingette H, Mailhe B, et al. Learning a probabilistic model for diffeomorphic registration[J]. IEEE transactions on medical imaging, 2019, 38(9): 2165-2176.
[8] Zhao S, Lau T F, Luo J, et al. Unsupervised 3d end-to-end medical image registration with volume tweening network[J]. IEEE Journal of Biomedical and Health Informatics, 2020, 24(5): 1394-1404.
[9] Vos B D, Berendsen F F, Viergever M A, et al. A deep learning framework for unsupervised affine and deformable image registration[J]. Medical image analysis, 2019, 52:128-143.
[10] Kim B, Kim D H, Park S H, et al. CycleMorph: cycle consistent unsupervised deformable image registration[J]. Medical Image Analysis, 2021, 71: 102036.
[11] Mok T C W, Chung A C S. Large deformation diffeomorphic image registration with Laplacian pyramid networks[C]// Medical Image Computing and Computer Assisted Intervention, 2020: 211-221.
[12] Kang M, Hu X, Huang W, et al. Dual-stream pyramid registration network[J]. Medical Image Analysis, 2022, 78: 102379.
[13] The Information eXtraction from Images (IXI) dataset. https://brain-development.org/ixi-dataset/. Accessed 19 Jan 2022.
[14] Avants B B, Epstein C L, Grossman M, et al. Symmetric diffeomorphic image registration with cross-correlation: evaluating automated labeling of elderly and neurodegenerative brain[J]. Medical Image Analysis, 2008, 12(1): 26-41.
[15] Beg, M F, Miller, M I, Trouv´E, A, Younes, L. Computing large deformation metric mappings via geodesic flows of diffeomorphisms[J]. International journal of computer vision ,2005, 61:139-157.
[16] Meng M, Bi L, Feng D, et al. Non-iterative Coarse-to-Fine Registration Based on Single-Pass Deep Cumulative Learning[C]//Proceeding of International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI). 2022: 88-97.