CoChat
Sign Up
Cited by 4Medical Image Segmentation TechniquesBrain Tumor Detection and ClassificationArtificial Intelligence in Healthcare and Education

The DOI matches its Crossref record; 3 of the 3 references it cites were found in a registry or an index.

Brain CT registration using hybrid supervised convolutional neural network

Open access
Yuan, Hongmei, Yang, Minglei, Qian, Shan, Wang, Wenxin, Jia, Xiaotian, Huang, Feng·2021·BioMedical Engineering OnLine
DOI 10.1186/s12938-021-00971-8PMC PMC8715595

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

Background: Image registration is an essential step in the automated interpretation of the brain computed tomography (CT) images of patients with acute cerebrovascular disease (ACVD). However, performing brain CT registration accurately and rapidly remains greatly challenging due to the large intersubject anatomical variations, low resolution of soft tissues, and heavy computation costs. To this end, the HSCN-Net, a hybrid supervised convolutional neural network, was developed for precise and fast brain CT registration. Method: HSCN-Net generated synthetic deformation fields using a simulator as one supervision for one reference–moving image pair to address the problem of lack of gold standards. Furthermore, the simulator was designed to generate multiscale affine and elastic deformation fields to overcome the registration challenge posed by large intersubject anatomical deformation. Finally, HSCN-Net adopted a hybrid loss function constituted by deformation field and image similarity to improve registration accuracy and generalization capability. In this work, 101 CT images of patients were collected for model construction (57), evaluation (14), and testing (30). HSCN-Net was compared with the classical Demons and VoxelMorph models. Qualitative analysis through the visual evaluation of critical brain tissues and quantitative analysis by determining the endpoint error (EPE) between the predicted sparse deformation vectors and gold-standard sparse deformation vectors, image normalized mutual information (NMI), and the Dice coefficient of the middle cerebral artery (MCA) blood supply area were carried out to assess model performance comprehensively. Results: HSCN-Net and Demons had a better visual spatial matching performance than VoxelMorph, and HSCN-Net was more competent for smooth and large intersubject deformations than Demons. The mean EPE of HSCN-Net (3.29 mm) was less than that of Demons (3.47 mm) and VoxelMorph (5.12 mm); the mean Dice of HSCN-Net was 0.96, which was higher than that of Demons (0.90) and VoxelMorph (0.87); and the mean NMI of HSCN-Net (0.83) was slightly lower than that of Demons (0.84), but higher than that of VoxelMorph (0.81). Moreover, the mean registration time of HSCN-Net (17.86 s) was shorter than that of VoxelMorph (18.53 s) and Demons (147.21 s). Conclusion: The proposed HSCN-Net could achieve accurate and rapid intersubject brain CT registration.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1186/s12938-021-00971-8

Cites

3 of 3 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Registration results of a typical brain CT image: a is the reference image; b is the moving image; and c–e are the registration results of HSCN-Net, Demons, and VoxelMorph algorithms, respectively; f–h are the local enlarged images corresponding to the same parts in c–e. All the three registration methods can achieve spatial matching. Compared with the ventricle in e, that in c and d had better spatial matching. The deformations in f and h were smoother than those in g and were highly consistent with the smooth characteristics of tissue deformation
Fig. 1 Registration results of a typical brain CT image: a is the reference image; b is the moving image; and c–e are the registration results of HSCN-Net, Demons, and VoxelMorph algorithms, respectively; f–h are the local enlarged images corresponding to the same parts in c–e. All the three registration methods can achieve spatial matching. Compared with the ventricle in e, that in c and d had better spatial matching. The deformations in f and h were smoother than those in g and were highly consistent with the smooth characteristics of tissue deformationCC BY · PubMed Central
Fig. 2 Framework of HSCN-Net. We used a simulator to generate synthetic deformation fields and images during training to address the lack of registration gold standards. Large deformations of appropriate proportions were added to the synthetic deformation fields to overcome the challenge of large deformation fields between the images to be registered. HSCN-Net adopts a hybrid supervised loss function that is constituted by the alignment accuracy of the deformation fields and image similarity to improve the registration accuracy and generalization capability of the model
Fig. 2 Framework of HSCN-Net. We used a simulator to generate synthetic deformation fields and images during training to address the lack of registration gold standards. Large deformations of appropriate proportions were added to the synthetic deformation fields to overcome the challenge of large deformation fields between the images to be registered. HSCN-Net adopts a hybrid supervised loss function that is constituted by the alignment accuracy of the deformation fields and image similarity to improve the registration accuracy and generalization capability of the modelCC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 3 references

Research works printed but not linked: 25

  1. Lee SH. Stroke revisited: diagnosis and treatment of ischemic stroke. In: Critical and medical management in acute stage of ischemic stroke (Chapter 14). 2017. p. 157–69. https://doi.org/10.1007/978-981-10-1424-6. · DOI
  2. Jung SM, Whangbo TK. Estimating aspect score for stroke from brain CT images based-on deep-learning. Adv Sci Technol. 2018;150:342–6. · DOI
  3. Mayer A, Zholkover A, Portnoy O, et al. Deformable registration of trans-rectal ultrasound (TRUS) and magnetic resonance imaging (MRI) for focal prostate brachytherapy. Int J Comput Assist Radiol Surg. 2016;11(6):1015–23. · DOI
  4. Broggi S, Scalco E, Belli ML, et al. A comparative evaluation of 3 different free-form deformable image registration and contour propagation methods for head and neck MRI: the case of parotid changes during radiotherapy. Technol Cancer Res Treat. 2017;16(3):373–81. · DOI
  5. Reaungamornrat S, Silva TD, Uneri A, et al. MIND demons: symmetric diffeomorphic deformable registration of MR and CT for image-guided spine surgery. IEEE Trans Med Imaging. 2016;35(11):2413–24. · DOI
  6. Krivov E, Pisov M, Belyaev M. MRI augmentation via elastic registration for brain lesions segmentation. In: International MICCAI brainlesion workshop. Cham: Springer; 2017. · DOI
  7. Liu X, Xing F, Yang C, et al. Symmetric-constrained irregular structure inpainting for brain MRI registration with tumor pathology. 2021. · DOI
  8. Uus A, Zhang T, Jackson LH, et al. Deformable slice-to-volume registration for motion correction of fetal body and placenta MRI. IEEE Trans Med Imaging. 2020;39(9):2750–9. · DOI
  9. Yu W, Tannast M, et al. Non-rigid free-form 2D–3D registration using a B-spline-based statistical deformation model. Pattern Recogn. 2017;63:689–99. · DOI
  10. Liu C, Zhang L, Wang L, et al. Multi-scale B-spline medical image registration based on pixel reconstruction. Intell Comput Appl. 2019;9(1):4.
  11. Wodzinski M, Skalski A, Ciepiela I, et al. Application of demons image registration algorithms in resected breast cancer lodge localization. In: 2017 signal processing: algorithms, architectures, arrangements, and applications (SPA). IEEE; 2017. · DOI
  12. Eppenhof K, Pluim J. Pulmonary CT registration through supervised learning with convolutional neural networks. IEEE Trans Med Imaging. 2019;38(5):1097–105. · DOI
  13. Dalca AV, Balakrishnan G, Guttag J, et al. Unsupervised learning for fast probabilistic diffeomorphic registration. 2018. · DOI
  14. Fan J, Cao X, Yap PT, et al. BIRNet: Brain image registration using dual-supervised fully convolutional networks. Med Image Anal. 2019;54:193–206. · DOI · Paper page
  15. Han R, Jones CK, Ketcha MD, et al. Deformable MR-CT image registration using an unsupervised end-to-end synthesis and registration network for endoscopic neurosurgery. In: Image-guided procedures, robotic interventions, and modeling. 2021. · DOI
  16. Kim B, Kim J, Lee J G, et al. Unsupervised deformable image registration using cycle-consistent CNN. 2019. · DOI
  17. Zhu W, Myronenko A, Xu Z, et al. NeurReg: neural registration and its application to image segmentation. In: 2020 IEEE winter conference on applications of computer vision (WACV). IEEE; 2020. · DOI
  18. Haskins G, Kruger U, Yan P. Deep learning in medical image registration: a survey. Mach Vis Appl. 2020;31(1–2):1–18.
  19. Mok T, Chung A. Large deformation diffeomorphic image registration with laplacian pyramid networks. 2020. · DOI
  20. Abualigah L, Diabat A, Mirjalili S, et al. The Arithmetic Optimization Algorithm. Comput Methods Appl Mech Eng. 2021;376:113609. · DOI
  21. Abualigah L, Yousri D, Elaziz MA, et al. Aquila optimizer: a novel meta-heuristic optimization algorithm. Comput Ind Eng. 2021;157:107250. · DOI
  22. Kingma D, Ba J. Adam: a method for stochastic optimization. 2014;6(12). arXiv:1412.6980. arXiv preprint · arXiv
  23. Abualigah L, Diabat A, Sumari P, Gandomi AH. A novel evolutionary arithmetic optimization algorithm for multilevel thresholding segmentation of covid-19 ct images. Processes. 2021;9(7):1155. https://doi.org/10.3390/pr9071155. · DOI
  24. Studholme C, Hill D, Hawkes DJ. An overlap invariant entropy measure of 3D medical image alignment. Pattern Recogn. 1999;32(1):71–86. · DOI
  25. Sudre C H, Li W, Vercauteren T, et al. Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentations. In: International workshop on deep learning in medical image analysis international workshop on multimodal learning for clinical decision support. 2017. · DOI · Paper page

Metadata

Authors
Yuan, H., Yang, M., Qian, S., Wang, W., Jia, X., & Huang, F.
Affiliations
  • Neusoft (China) (CN)
  • Neusoft Medical Systems Co., Ltd. (China) (CN)
Corresponding author
Yang, Minglei
ORCID
Venue
BioMedical Engineering OnLine
Publisher
Springer Science and Business Media LLC
Published
2021-12
Volume
20
Issue
1
Pages
131-131
ISSN
1475-925X (ISSN-L)
DOI
10.1186/s12938-021-00971-8
PMC
PMC8715595
OpenAlex
W4225781432
Cited by
4
References
3
Citations by year
2026: 1 · 2024: 1 · 2023: 1 · 2022: 1
FWCI
0.27
Type
journal article
Language
en
Open access
Licence
Locations
  • BioMedical Engineering OnLine (publishedVersion, cc-by)
  • Research Square (acceptedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
shenyang science and technology plan fund (20-201-4-10)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Medical Image Segmentation Techniques
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Brain Tumor Detection and Classification
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
Keywords
image registration, hybrid loss function, acute cerebrovascular disease, demons, VoxelMorph, endpoint error, registration speed, brain CT images
MeSH terms
Algorithms, Brain, Humans, Image Processing, Computer-Assisted, Tomography, X-Ray Computed, Neural Networks, Computer

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Brain CT registration using hybrid supervised convolutional neural network. https://app.cochat.ai/papers/10.1186/s12938-021-00971-8
The paper
Yuan, H., Yang, M., Qian, S., Wang, W., Jia, X., & Huang, F. (2021). Brain CT registration using hybrid supervised convolutional neural network. BioMedical Engineering OnLine. https://doi.org/10.1186/s12938-021-00971-8