Skip to content
2000
Volume 19, Issue 1
  • ISSN: 2666-2558
  • E-ISSN: 2666-2566

Abstract

Introduction

Deformable image registration is an essential task in medical image analysis. The UNet model, or the model with the U-shaped structure, has been popularly proposed in deep learning-based registration methods. However, they easily lose the important similarity information in the up-sampling stage, and these methods usually ignore the inherent inverse consistency of the transformation between a pair of images. Furthermore, the traditional smoothing constraints used in the existing methods can only partially ensure the folding of the deformation field.

Methods

An inverse consistent deformable medical image registration network (ICSANet) based on the inverse consistency constraint and the similarity-based local attention is developed. A new UNet network is constructed by introducing similarity-based local attention to focus on the spatial correspondence in the high-similarity space. A novel inverse consistency constraint is proposed, and the objective function of the new form is presented with the combination of the traditional constraint conditions.

Experiment

The performance of the proposed method is compared with the typical registration models, such as the VoxelMorph, PVT, nnFormer, and TransMorph-diff model, on the brain IXI and OASIS datasets.

Results

Experimental results on the brain MRI datasets show that the images can be deformed symmetrically until two distorted images are well matched. The quantitative comparison and visual analysis indicate that the proposed method performs better, and the Dice index can be improved by at least 12% with only 10% parameters.

Conclusion

This paper presents a new medical image registration network, ICSANet. By introducing a similarity attention gate, it accurately captures high-similarity spatial correspondences between source and target images, resulting in better registration performance.

© 2026 Bentham Science Publishers
Loading

Article metrics loading...

/content/journals/rascs/10.2174/0126662558338439241029180737
2024-12-09
2025-12-17

  • From This Site

    /content/journals/rascs/10.2174/0126662558338439241029180737
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

References

  1. GaserC. Structural MRI: Morphometry.Neuroeconomics2016399409
     [Google Scholar]
  2. CaiX. ZhangJ. LuJ. YiL. HanZ. ZhangS. YangX. LiuG. N ‐Aryl amides as chemical exchange saturation transfer magnetic resonance imaging contrast agents.Chemistry20202651117051170910.1002/chem.20200241532639618
     [Google Scholar]
  3. ZhangJ. HanZ. LuJ. LiY. LiaoX. van ZijlP.C. YangX. LiuG. Triazoles as T2 ‐Exchange magnetic resonance imaging contrast agents for the detection of nitrilase activity.Chemistry20182456150131501810.1002/chem.20180266329989227
     [Google Scholar]
  4. LianC. ZhangJ. LiuM. ZongX. HungS.C. LinW. ShenD. Multi-channel multi-scale fully convolutional network for 3D perivascular spaces segmentation in 7T MR images.Med. Image Anal.20184610611710.1016/j.media.2018.02.00929518675
     [Google Scholar]
  5. ChatterjeeS. BajajH. SiddiqueeI.H. SubbarayappaN.B. SimonS. ShashidharS.B. SpeckO. NürnbergerA. MICDIR: Multi-scale inverse-consistent deformable image registration using UNetMSS with self-constructing graph latent.Comput. Med. Imaging Graph.202310810226710.1016/j.compmedimag.2023.10226737506427
     [Google Scholar]
  6. MaX. CuiH. LiS. YangY. XiaY. Deformable medical image registration with global–local transformation network and region similarity constraint.Comput. Med. Imaging Graph.202310810226310.1016/j.compmedimag.2023.10226337487363
     [Google Scholar]
  7. ZhangJ. WangJ. WangX. FengD. The adaptive FEM elastic model for medical image registration.Phys. Med. Biol.20145919711810.1088/0031‑9155/59/1/9724334618
     [Google Scholar]
  8. AvantsB. EpsteinC. GrossmanM. GeeJ. Symmetric diffeomorphic image registration with cross-correlation: Evaluating automated labeling of elderly and neurodegenerative brain.Med. Image Anal.2008121264110.1016/j.media.2007.06.00417659998
     [Google Scholar]
  9. ZhangW. ZhaoY. Hierarchical registration of brain images based on B-splines and Laplacian commutators.Optik (Stuttg.)202124116702210.1016/j.ijleo.2021.167022
     [Google Scholar]
  10. SimonovskyM. Gutiérrez-BeckerB. MateusD. A deep metric for multimodal registration19th International ConferenceAthens20161018
     [Google Scholar]
  11. LiaoR. MiaoS. De TournemireP. GrbicS. KamenA. MansiT. ComaniciuD. An artificial agent for robust image registration.Proc. Conf. AAAI Artif. Intell.201731110.1609/aaai.v31i1.11230
     [Google Scholar]
  12. BlendowskiM. HansenL. HeinrichM.P. Weakly-supervised learning of multi-modal features for regularised iterative descent in 3D image registration.Med. Image Anal.20216710182210.1016/j.media.2020.10182233166774
     [Google Scholar]
  13. WuM. HeX. LiF. ZhuJ. WangS. BursteinP.D. Weakly supervised volumetric prostate registration for MRI-TRUS image driven by signed distance map.Comput. Biol. Med.202316310715010.1016/j.compbiomed.2023.10715037321103
     [Google Scholar]
  14. CaoX. YangJ. ZhangJ. WangQ. YapP.T. ShenD. Deformable image registration using a cue-aware deep regression network.IEEE Trans. Biomed. Eng.20186591900191110.1109/TBME.2018.282282629993391
     [Google Scholar]
  15. MiaoS. WangZ.J. ZhengY. LiaoR. Real-time 2D/3D registration via CNN regressionIEEE 13th International Symposium on Biomedical Imaging (ISBI)Prague, Czech Republic 2016, pp. 1430-143410.1109/ISBI.2016.7493536
     [Google Scholar]
  16. FanJ. CaoX. YapP.T. ShenD. BIRNet: Brain image registration using dual-supervised fully convolutional networks.Med. Image Anal.20195419320610.1016/j.media.2019.03.00630939419
     [Google Scholar]
  17. LuoX. DuB. GuiP. ZhangD. HuW. A Hunger Games Search algorithm with opposition-based learning for solving multimodal medical image registration.Neurocomputing202354012620410.1016/j.neucom.2023.03.065
     [Google Scholar]
  18. ZhengZ. CaoW. LianD. LuoY. Feature self-calibration network with global-local training strategy for multi-region deformable medical image registration.Neural Comput. Appl.20223419171751719110.1007/s00521‑022‑07365‑4
     [Google Scholar]
  19. MinZ. Non-rigid Medical Image Registration using Physics-informed Neural NetworksInternational Conference on Information Processing in Medical Imaging San Carlos de BarilocheArgentina2023182310.1007/978‑3‑031‑34048‑2_46
     [Google Scholar]
  20. BalakrishnanG. ZhaoA. SabuncuM.R. An unsupervised learning model for deformable medical image registrationProceedings of the IEEE Conference on Computer Vision and Pattern Recognition Salt Lake CityUT, USA20189252926010.1109/CVPR.2018.00964
     [Google Scholar]
  21. BalakrishnanG. ZhaoA. SabuncuM.R. GuttagJ. DalcaA.V. VoxelMorph: A learning framework for deformable medical image registration.IEEE Trans. Med. Imaging20193881788180010.1109/TMI.2019.289753830716034
     [Google Scholar]
  22. ZhangJ. Inverse-consistent deep networks for unsupervised deformable image registrationarXiv2018
     [Google Scholar]
  23. VercauterenT. PennecX. PerchantA. AyacheN. Diffeomorphic demons: Efficient non-parametric image registration.Neuroimage2009451, Suppl.S61S7210.1016/j.neuroimage.2008.10.04019041946
     [Google Scholar]
  24. ZhaoS. LauT. LuoJ. ChangE.I.C. XuY. Unsupervised 3D end-to-end medical image registration with volume tweening network.IEEE J. Biomed. Health Inform.20202451394140410.1109/JBHI.2019.295102431689224
     [Google Scholar]
  25. CaiG. LiuH. ZouW. HuN. WangJ. Registration of 3D medical images based on unsupervised cooperative cascade of deep networks.Biomed. Signal Process. Control20238210459410.1016/j.bspc.2023.104594
     [Google Scholar]
  26. ValanarasuJ.M.J. OzaP. HacihalilogluI. PatelV.M. Medical transformer: Gated axial-attention for medical image segmentation24th International Conference StrasbourgFrance2021364610.1007/978‑3‑030‑87193‑2_4
     [Google Scholar]
  27. ChristensenG.E. JohnsonH.J. Consistent image registration.IEEE Trans. Med. Imaging200120756858210.1109/42.93274211465464
     [Google Scholar]
  28. LeowA. Inverse consistent mapping in 3D deformable image registration: its construction and statistical properties19th international conference on Information Processing in Medical Imaging (IPMI’05)Springer-Verlag, Berlin, Heidelberg, 2005, 493–50310.1007/11505730_41
     [Google Scholar]
  29. HeJ. ChristensenG.E. Large deformation inverse consistent elastic image registrationBiennial International Conference on Information Processing in Medical ImagingBerlin, Heidelberg, 2003, pp. 438-44910.1007/978‑3‑540‑45087‑0_37
     [Google Scholar]
  30. ColeJ.H. PoudelR.P.K. TsagkrasoulisD. CaanM.W.A. StevesC. SpectorT.D. MontanaG. Spector, Giovanni Montana, Predicting brain age with deep learning from raw imaging data results in a reliable and heritable biomarker.Neuroimage2017163115124
     [Google Scholar]
  31. MarcusD.S. WangT.H. ParkerJ. CsernanskyJ.G. MorrisJ.C. BucknerR.L. Open Access Series of Imaging Studies (OASIS): Cross-sectional MRI data in young, middle aged, nondemented, and demented older adults.J. Cogn. Neurosci.20071991498150710.1162/jocn.2007.19.9.149817714011
     [Google Scholar]
  32. MiaoS. WangZ.J. LiaoR. A CNN regression approach for real-time 2D/3D registration.IEEE Trans. Med. Imaging20163551352136310.1109/TMI.2016.252180026829785
     [Google Scholar]
  33. AvantsB.B. TustisonN.J. WuJ. CookP.A. GeeJ.C. An open source multivariate framework for n-tissue segmentation with evaluation on public data.Neuroinformatics20119438140010.1007/s12021‑011‑9109‑y21373993
     [Google Scholar]
  34. De VosB.D. BerendsenF.F. ViergeverM.A. StaringM. IšgumI. End-to-end unsupervised deformable image registration with a convolutional neural networkThird International Workshop, DLMIA 2017, and 7th International Workshop, ML-CDS Cardoso2017, pp. 204-212.10.1007/978‑3‑319‑67558‑9_24
     [Google Scholar]
  35. ZhouH-Y. GuoJ. ZhangY. YuL. WangL. YuY. nnformer: Interleaved transformer for volumetric segmentationarXiv20211
     [Google Scholar]
  36. WangW. Pyramid vision transformer: A versatile backbone for dense prediction without convolutionsProceedings of the IEEE/CVF International Conference on Computer Vision Canada202154855810.1109/ICCV48922.2021.00061
     [Google Scholar]
  37. ChenJ. HeY. FreyE.C. LiY. DuY. Vit-v-net: Vision transformer for unsupervised volumetric medical image registrationarXiv2021
     [Google Scholar]
  38. XieY. ZhangJ. ShenC. XiaY. Cotr: Efficiently bridging cnn and transformer for 3d medical image segmentation24th International ConferenceFrance202117118010.1007/978‑3‑030‑87199‑4_16
     [Google Scholar]
  39. ModatM. RidgwayG.R. TaylorZ.A. LehmannM. BarnesJ. HawkesD.J. FoxN.C. OurselinS. Fast free-form deformation using graphics processing units.Comput. Methods Programs Biomed.201098327828410.1016/j.cmpb.2009.09.00219818524
     [Google Scholar]
  40. BegM.F. MillerM.I. TrouvéA. YounesL. Computing large deformation metric mappings via geodesic flows of diffeomorphisms.Int. J. Comput. Vis.200561213915710.1023/B:VISI.0000043755.93987.aa
     [Google Scholar]
  41. HeinrichM.P. MaierO. HandelsH. Multi-modal multi-atlas segmentation using discrete optimisation and self-similaritiesIEEE International Symposium on Biomedical ImagingNew York, NY201527
     [Google Scholar]
  42. QiuH. QinC. SchuhA. Learning diffeomorphic and modality-invariant registration using b-splinesFourth Conference on Medical Imaging with Deep Learning, PMLR 012021645664
     [Google Scholar]
  43. KimB. KimD.H. ParkS.H. KimJ. LeeJ.G. YeJ.C. CycleMorph: Cycle consistent unsupervised deformable image registration.Med. Image Anal.20217110203610.1016/j.media.2021.10203633827038
     [Google Scholar]
  44. ChenJ. FreyE.C. HeY. SegarsW.P. LiY. DuY. TransMorph: Transformer for unsupervised medical image registration.Med. Image Anal.20228210261510.1016/j.media.2022.10261536156420
     [Google Scholar]
  45. MattesD. HaynorD.R. VesselleH. LewellenT.K. EubankW. PET-CT image registration in the chest using free-form deformations.IEEE Trans. Med. Imaging200322112012810.1109/TMI.2003.80907212703765
     [Google Scholar]
  46. StaringM. van der HeideU.A. KleinS. ViergeverM.A. PluimJ. Registration of cervical MRI using multifeature mutual information.IEEE Trans. Med. Imaging20092891412142110.1109/TMI.2009.201656019278929
     [Google Scholar]
  47. MatsuoH. NishioM. NogamiM. ZengF. KurimotoT. KaushikS. WiesingerF. KonoA.K. MurakamiT. Unsupervised-learning-based method for chest MRI–CT transformation using structure constrained unsupervised generative attention networks.Sci. Rep.20221211109010.1038/s41598‑022‑14677‑x35773366
     [Google Scholar]
  48. PaulyO. PadoyN. PoppertH. EspositoL. NavabN. Wavelet energy map: A robust support for multi-modal registration of medical imagesProceedings of the IEEE Conference on Computer Vision and Pattern Recognition MiamiFL, USA, 2009, pp. 2184-2191.10.1109/CVPR.2009.5206528
     [Google Scholar]
  49. WachingerC. NavabN. Entropy and Laplacian images: Structural representations for multi-modal registration.Med. Image Anal.201216111710.1016/j.media.2011.03.00121632274
     [Google Scholar]
/content/journals/rascs/10.2174/0126662558338439241029180737
Loading
The Inverse-Consistent Deformable MRI Registration Method Based on the Improved UNet Model and Similarity Attention
Recent Advances in Computer Science and Communications 19, E26662558338439 (2026); https://doi.org/10.2174/0126662558338439241029180737
/content/journals/rascs/10.2174/0126662558338439241029180737
/content/journals/rascs/10.2174/0126662558338439241029180737
Loading

Data & Media loading...


  • Article Type:     Research Article
Keyword(s): image registration; inverse-consistency; Medical image; registration performance; similarity attention; UNet

Most Cited Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error
Please enter a valid_number test