Groupwise registration via graph shrinkage on the image manifold

Shihui Ying, Guorong Wu, Qian Wang, Dinggang Shen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

13 Citations (Scopus)

Abstract

Recently, group wise registration has been investigated for simultaneous alignment of all images without selecting any individual image as the template, thus avoiding the potential bias in image registration. However, none of current group wise registration method fully utilizes the image distribution to guide the registration. Thus, the registration performance usually suffers from large inter-subject variations across individual images. To solve this issue, we propose a novel group wise registration algorithm for large population dataset, guided by the image distribution on the manifold. Specifically, we first use a graph to model the distribution of all image data sitting on the image manifold, with each node representing an image and each edge representing the geodesic pathway between two nodes (or images). Then, the procedure of warping all images to their population center turns to the dynamic shrinking of the graph nodes along their graph edges until all graph nodes become close to each other. Thus, the topology of image distribution on the image manifold is always preserved during the group wise registration. More importantly, by modeling the distribution of all images via a graph, we can potentially reduce registration error since every time each image is warped only according to its nearby images with similar structures in the graph. We have evaluated our proposed group wise registration method on both synthetic and real datasets, with comparison to the two state-of-the-art group wise registration methods. All experimental results show that our proposed method achieves the best performance in terms of registration accuracy and robustness.

Original languageEnglish
Title of host publicationProceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
Pages2323-2330
Number of pages8
DOIs
Publication statusPublished - 2013
Externally publishedYes
Event26th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2013 - Portland, OR, United States
Duration: 2013 Jun 232013 Jun 28

Other

Other26th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2013
CountryUnited States
CityPortland, OR
Period13/6/2313/6/28

Fingerprint

Image registration
Topology

Keywords

  • diffeomorphism
  • graph shrinking
  • image manifold
  • Unbiased groupwise registration

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition

Cite this

Ying, S., Wu, G., Wang, Q., & Shen, D. (2013). Groupwise registration via graph shrinkage on the image manifold. In Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (pp. 2323-2330). [6619145] https://doi.org/10.1109/CVPR.2013.301

Groupwise registration via graph shrinkage on the image manifold. / Ying, Shihui; Wu, Guorong; Wang, Qian; Shen, Dinggang.

Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. 2013. p. 2323-2330 6619145.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ying, S, Wu, G, Wang, Q & Shen, D 2013, Groupwise registration via graph shrinkage on the image manifold. in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition., 6619145, pp. 2323-2330, 26th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2013, Portland, OR, United States, 13/6/23. https://doi.org/10.1109/CVPR.2013.301
Ying S, Wu G, Wang Q, Shen D. Groupwise registration via graph shrinkage on the image manifold. In Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. 2013. p. 2323-2330. 6619145 https://doi.org/10.1109/CVPR.2013.301
Ying, Shihui ; Wu, Guorong ; Wang, Qian ; Shen, Dinggang. / Groupwise registration via graph shrinkage on the image manifold. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. 2013. pp. 2323-2330
@inproceedings{97dd72b7991e4c559996803d38ff37c7,
title = "Groupwise registration via graph shrinkage on the image manifold",
abstract = "Recently, group wise registration has been investigated for simultaneous alignment of all images without selecting any individual image as the template, thus avoiding the potential bias in image registration. However, none of current group wise registration method fully utilizes the image distribution to guide the registration. Thus, the registration performance usually suffers from large inter-subject variations across individual images. To solve this issue, we propose a novel group wise registration algorithm for large population dataset, guided by the image distribution on the manifold. Specifically, we first use a graph to model the distribution of all image data sitting on the image manifold, with each node representing an image and each edge representing the geodesic pathway between two nodes (or images). Then, the procedure of warping all images to their population center turns to the dynamic shrinking of the graph nodes along their graph edges until all graph nodes become close to each other. Thus, the topology of image distribution on the image manifold is always preserved during the group wise registration. More importantly, by modeling the distribution of all images via a graph, we can potentially reduce registration error since every time each image is warped only according to its nearby images with similar structures in the graph. We have evaluated our proposed group wise registration method on both synthetic and real datasets, with comparison to the two state-of-the-art group wise registration methods. All experimental results show that our proposed method achieves the best performance in terms of registration accuracy and robustness.",
keywords = "diffeomorphism, graph shrinking, image manifold, Unbiased groupwise registration",
author = "Shihui Ying and Guorong Wu and Qian Wang and Dinggang Shen",
year = "2013",
doi = "10.1109/CVPR.2013.301",
language = "English",
pages = "2323--2330",
booktitle = "Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition",

}

TY - GEN

T1 - Groupwise registration via graph shrinkage on the image manifold

AU - Ying, Shihui

AU - Wu, Guorong

AU - Wang, Qian

AU - Shen, Dinggang

PY - 2013

Y1 - 2013

N2 - Recently, group wise registration has been investigated for simultaneous alignment of all images without selecting any individual image as the template, thus avoiding the potential bias in image registration. However, none of current group wise registration method fully utilizes the image distribution to guide the registration. Thus, the registration performance usually suffers from large inter-subject variations across individual images. To solve this issue, we propose a novel group wise registration algorithm for large population dataset, guided by the image distribution on the manifold. Specifically, we first use a graph to model the distribution of all image data sitting on the image manifold, with each node representing an image and each edge representing the geodesic pathway between two nodes (or images). Then, the procedure of warping all images to their population center turns to the dynamic shrinking of the graph nodes along their graph edges until all graph nodes become close to each other. Thus, the topology of image distribution on the image manifold is always preserved during the group wise registration. More importantly, by modeling the distribution of all images via a graph, we can potentially reduce registration error since every time each image is warped only according to its nearby images with similar structures in the graph. We have evaluated our proposed group wise registration method on both synthetic and real datasets, with comparison to the two state-of-the-art group wise registration methods. All experimental results show that our proposed method achieves the best performance in terms of registration accuracy and robustness.

AB - Recently, group wise registration has been investigated for simultaneous alignment of all images without selecting any individual image as the template, thus avoiding the potential bias in image registration. However, none of current group wise registration method fully utilizes the image distribution to guide the registration. Thus, the registration performance usually suffers from large inter-subject variations across individual images. To solve this issue, we propose a novel group wise registration algorithm for large population dataset, guided by the image distribution on the manifold. Specifically, we first use a graph to model the distribution of all image data sitting on the image manifold, with each node representing an image and each edge representing the geodesic pathway between two nodes (or images). Then, the procedure of warping all images to their population center turns to the dynamic shrinking of the graph nodes along their graph edges until all graph nodes become close to each other. Thus, the topology of image distribution on the image manifold is always preserved during the group wise registration. More importantly, by modeling the distribution of all images via a graph, we can potentially reduce registration error since every time each image is warped only according to its nearby images with similar structures in the graph. We have evaluated our proposed group wise registration method on both synthetic and real datasets, with comparison to the two state-of-the-art group wise registration methods. All experimental results show that our proposed method achieves the best performance in terms of registration accuracy and robustness.

KW - diffeomorphism

KW - graph shrinking

KW - image manifold

KW - Unbiased groupwise registration

UR - http://www.scopus.com/inward/record.url?scp=84887324660&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84887324660&partnerID=8YFLogxK

U2 - 10.1109/CVPR.2013.301

DO - 10.1109/CVPR.2013.301

M3 - Conference contribution

AN - SCOPUS:84887324660

SP - 2323

EP - 2330

BT - Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition

ER -