Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
97 tokens/sec
GPT-4o
53 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
5 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

ModeTv2: GPU-accelerated Motion Decomposition Transformer for Pairwise Optimization in Medical Image Registration (2403.16526v1)

Published 25 Mar 2024 in cs.CV

Abstract: Deformable image registration plays a crucial role in medical imaging, aiding in disease diagnosis and image-guided interventions. Traditional iterative methods are slow, while deep learning (DL) accelerates solutions but faces usability and precision challenges. This study introduces a pyramid network with the enhanced motion decomposition Transformer (ModeTv2) operator, showcasing superior pairwise optimization (PO) akin to traditional methods. We re-implement ModeT operator with CUDA extensions to enhance its computational efficiency. We further propose RegHead module which refines deformation fields, improves the realism of deformation and reduces parameters. By adopting the PO, the proposed network balances accuracy, efficiency, and generalizability. Extensive experiments on two public brain MRI datasets and one abdominal CT dataset demonstrate the network's suitability for PO, providing a DL model with enhanced usability and interpretability. The code is publicly available.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (76)
  1. Symmetric diffeomorphic image registration with cross-correlation: Evaluating automated labeling of elderly and neurodegenerative brain. Medical Image Analysis 12, 26–41.
  2. Advanced normalization tools (ANTS). Insight j 2, 1–35.
  3. Layer normalization. arXiv preprint arXiv:1607.06450 .
  4. Multiresolution elastic matching. Computer Vision, Graphics, and Image Processing 46, 1–21.
  5. VoxelMorph: A learning framework for deformable medical image registration. IEEE Transactions on Medical Imaging 38, 1788–1800.
  6. Computing large deformation metric mappings via geodesic flows of diffeomorphisms. International Journal of Computer Vision 61, 139–157.
  7. Edge-aware pyramidal deformable network for unsupervised registration of brain MR images. Frontiers in Neuroscience 14, 620235.
  8. Transmorph: Transformer for unsupervised medical image registration. Medical Image Analysis 82, 102615.
  9. Vit-V-Net: Vision transformer for unsupervised volumetric medical image registration. arXiv preprint arXiv:2104.06468 .
  10. Deformable cross-attention transformer for medical image registration, in: Machine Learning in Medical Imaging, pp. 115–125.
  11. Deformer: Towards displacement field learning for unsupervised medical image registration, in: Internatioal Conference on Medical Image Computing and Computer Assisted Intervention, pp. 141–151.
  12. DeepLab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Transactions on Pattern Analysis and Machine Intelligence 40, 834–848.
  13. TransMatch: A transformer-based multilevel dual-stream feature matching network for unsupervised deformable image registration. IEEE Transactions on Medical Imaging 43, 15–27.
  14. Empirical evaluation of gated recurrent neural networks on sequence modeling, in: NIPS 2014 Workshop on Deep Learning.
  15. Unsupervised learning of probabilistic diffeomorphic registration for images and surfaces. Medical Image Analysis 57, 226–236.
  16. Measures of the amount of ecologic association between species. Ecology 26, 297–302.
  17. An image is worth 16x16 words: Transformers for image recognition at scale, in: International Conference on Learning Representations.
  18. One-shot learning for deformable medical image registration and periodic motion tracking. IEEE Transactions on Medical Imaging 39, 2506–2517.
  19. On the adaptability of unsupervised CNN-based deformable image registration to unseen image domains, in: Machine Learning in Medical Imaging, Springer. pp. 294–302.
  20. Slice-to-volume medical image registration: A survey. Medical Image Analysis 39, 101–123.
  21. FreeSurfer. NeuroImage 62, 774–781.
  22. Deep learning in medical image registration: a review. Physics in Medicine & Biology 65, 20TR01.
  23. Deep learning in medical image registration: a survey. Machine Vision and Applications 31, 1–18.
  24. MIND: Modality independent neighbourhood descriptor for multi-modal deformable registration. Medical Image Analysis 16, 1423–1435.
  25. Deformable image registration by combining uncertainty estimates from supervoxel belief propagation. Medical Image Analysis 27, 57–71.
  26. mlVIRNET: Multilevel variational image registration network, in: International Conference on Medical Image Computing and Computer Assisted Intervention, Springer. pp. 257–265.
  27. Recursive decomposition network for deformable image registration. IEEE Journal of Biomedical and Health Informatics 26, 5130–5141.
  28. Squeeze-and-excitation networks, in: International Conference on Computer Vision and Pattern Recognition, pp. 7132–7141.
  29. Dual-stream pyramid registration network, in: Internatioal Conference on Medical Image Computing and Computer Assisted Intervention, pp. 382–390.
  30. Comparing images using the hausdorff distance. IEEE Transactions on Pattern Analysis and Machine Intelligence 15, 850–863.
  31. One shot PACS: Patient specific anatomic context and shape prior aware recurrent registration-segmentation of longitudinal thoracic cone beam CTs. IEEE Transactions on Medical Imaging 41, 2021–2032.
  32. Dual-stream pyramid registration network. Medical Image Analysis 78, 102379.
  33. Anatomically constrained and attention-guided deep feature fusion for joint segmentation and deformable medical image registration. Medical Image Analysis 88, 102811.
  34. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 .
  35. 101 labeled brain images and a consistent human cortical labeling protocol. Frontiers in Neuroscience 6, 171.
  36. Elastix: a toolbox for intensity-based medical image registration. IEEE Transactions on Medical Imaging 29, 196–205.
  37. Normalized mutual information based registration using k-means clustering and shading correction. Medical Image Analysis 10, 432–439.
  38. Dual attention network for unsupervised medical image registration based on voxelmorph. Scientific Reports 12, 16250.
  39. Coordinate translator for learning deformable medical image registration, in: Multiscale Multimodal Medical Imaging, pp. 98–109.
  40. Swin Transformer: Hierarchical vision transformer using shifted windows, in: International Conference on Computer Vision, pp. 10012–10022.
  41. Joint progressive and coarse-to-fine registration of brain MRI via deformation field integration and non-rigid feature fusion. IEEE Transactions on Medical Imaging 41, 2788–2802.
  42. PIViT: Large deformation image registration with pyramid-iterative vision transformer, in: International Conference on Medical Image Computing and Computer Assisted Intervention, Springer. pp. 602–612.
  43. Non-iterative coarse-to-fine registration based on single-pass deep cumulative learning, in: International Conference on Medical Image Computing and Computer Assisted Intervention, Springer. pp. 88–97.
  44. Non-iterative coarse-to-fine transformer networks for joint affine and deformable image registration, in: International Conference on Medical Image Computing and Computer Assisted Intervention, Springer. pp. 750–760.
  45. Large deformation diffeomorphic image registration with laplacian pyramid networks, in: Internatioal Conference on Medical Image Computing and Computer Assisted Intervention. Springer, pp. 211–221.
  46. Deformable medical image registration under distribution shifts with neural instance optimization, in: International Workshop on Machine Learning in Medical Imaging, Springer. pp. 126–136.
  47. Salient deformable network for abdominal multiorgan registration. Medical Physics 49, 5953–5963.
  48. Application of normalized cross correlation to image registration. International Journal of Research in Engineering and Technology 3, 12–16.
  49. U-Net: Convolutional networks for biomedical image segmentation, in: Internatioal Conference on Medical Image Computing and Computer Assisted Intervention, pp. 234–241.
  50. Diffeomorphic registration using B-splines, in: International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), Springer. pp. 702–709.
  51. Nonrigid registration using free-form deformations: application to breast MR images. IEEE Transactions on Medical Imaging 18, 712–721.
  52. Recurrent registration neural networks for deformable image registration. Advances in Neural Information Processing Systems 32.
  53. Construction of a 3D probabilistic atlas of human cortical structures. NeuroImage 39, 1064–1080.
  54. Xmorpher: Full transformer for deformable medical image registration via cross attention, in: Internatioal Conference on Medical Image Computing and Computer Assisted Intervention, pp. 217–226.
  55. Convolutional LSTM network: A machine learning approach for precipitation nowcasting. Advances in Neural Information Processing Systems 28.
  56. Cross-modal attention for multi-modal image registration. Medical Image Analysis 82, 102612.
  57. Cross-modal attention for MRI and ultrasound volume registration, in: Internatioal Conference on Medical Image Computing and Computer Assisted Intervention, pp. 66–75.
  58. Deformable medical image registration: A survey. IEEE Transactions on Medical Imaging 32, 1153–1190.
  59. Metrics for evaluating 3D medical image segmentation: analysis, selection, and tool. BMC Medical Imaging 15, 1–28.
  60. Image matching as a diffusion process: an analogy with Maxwell’s demons. Medical Image Analysis 2, 243–260.
  61. Open-source platforms for navigated image-guided interventions. Medical Image Analysis 33, 181–186.
  62. Attention is all you need, in: Advances in Neural Information Processing Systems.
  63. Diffeomorphic demons: Efficient non-parametric image registration. NeuroImage 45, S61–S72.
  64. ModeT: Learning deformable image registration via motion decomposition transformer, in: Internatioal Conference on Medical Image Computing and Computer Assisted Intervention, pp. 740–749.
  65. Recursive deformable pyramid network for unsupervised medical image registration. IEEE Transactions on Medical Imaging .
  66. Robust image registration using log-polar transform, in: International Conference on Image Processing, pp. 493–496.
  67. CBAM: Convolutional block attention module, in: European Conference on Computer Vision, pp. 3–19.
  68. Cross-modality image registration using a training-time privileged third modality. IEEE Transactions on Medical Imaging 41, 3421–3431.
  69. Fast normalized cross-correlation. Circuits, Systems and Signal Processing 28, 819–843.
  70. Recursive cascaded networks for unsupervised medical image registration, in: International Conference on Computer Vision, pp. 10600–10610.
  71. Unsupervised 3D end-to-end medical image registration with volume tweening network. IEEE Journal of Biomedical and Health Informatics 24, 1394–1404.
  72. Residual aligner-based network (ran): Motion-separable structure for coarse-to-fine discontinuous deformable registration. Medical Image Analysis 91, 103038.
  73. Recursive deformable image registration network with mutual attention, in: Medical Image Understanding and Analysis, pp. 75–86.
  74. Self-distilled hierarchical network for unsupervised deformable image registration. IEEE Transactions on Medical Imaging 42, 2162–2175.
  75. Similarity attention-based CNN for robust 3D medical image registration. Biomedical Signal Processing and Control 81, 104403.
  76. Swin-voxelmorph: A symmetric unsupervised learning model for deformable medical image registration using swin transformer, in: Internatioal Conference on Medical Image Computing and Computer Assisted Intervention, pp. 78–87.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (4)
  1. Haiqiao Wang (6 papers)
  2. Zhuoyuan Wang (16 papers)
  3. Dong Ni (90 papers)
  4. Yi Wang (1038 papers)
Citations (2)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now