Globally optimal deformable registration on a minimum spanning tree using dense displacement sampling

Mattias P. Heinrich, Mark Jenkinson, Sir Michael Brady, Julia A. Schnabel

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

26 Citations (Scopus)

Abstract

Deformable image registration poses a highly non-convex optimisation problem. Conventionally, medical image registration techniques rely on continuous optimisation, which is prone to local minima. Recent advances in the mathematics and new programming methods enable these disadvantages to be overcome using discrete optimisation. In this paper, we present a new technique deeds, which employs a discrete dense displacement sampling for the deformable registration of high resolution CT volumes. The image grid is represented as a minimum spanning tree. Given these constraints a global optimum of the cost function can be found efficiently using dynamic programming, which enforces the smoothness of the deformations. Experimental results demonstrate the advantages of deeds: the registration error for the challenging registration of inhale and exhale pulmonary CT scans is significantly lower than for two state-of-the-art registration techniques, especially in the presence of large deformations and sliding motion at lung surfaces.

LanguageEnglish
Title of host publicationMedical Image Computing and Computer-Assisted Intervention, MICCAI2012 - 15th International Conference, Proceedings
EditorsPolina Golland, Nicholas Ayache, Herve Delingette, Kensaku Mori
PublisherSpringer Verlag
Pages115-122
Number of pages8
ISBN (Print)9783642334535
Publication statusPublished - 1 Jan 2012
Externally publishedYes
Event15th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2012 - Nice, France
Duration: 1 Oct 20125 Oct 2012

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7512 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other15th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2012
CountryFrance
CityNice
Period1/10/125/10/12

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Heinrich, M. P., Jenkinson, M., Brady, S. M., & Schnabel, J. A. (2012). Globally optimal deformable registration on a minimum spanning tree using dense displacement sampling. In P. Golland, N. Ayache, H. Delingette, & K. Mori (Eds.), Medical Image Computing and Computer-Assisted Intervention, MICCAI2012 - 15th International Conference, Proceedings (pp. 115-122). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7512 LNCS). Springer Verlag.
Heinrich, Mattias P. ; Jenkinson, Mark ; Brady, Sir Michael ; Schnabel, Julia A. / Globally optimal deformable registration on a minimum spanning tree using dense displacement sampling. Medical Image Computing and Computer-Assisted Intervention, MICCAI2012 - 15th International Conference, Proceedings. editor / Polina Golland ; Nicholas Ayache ; Herve Delingette ; Kensaku Mori. Springer Verlag, 2012. pp. 115-122 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{cd6b341b2b6b4497a646270e9c8b1fcd,
title = "Globally optimal deformable registration on a minimum spanning tree using dense displacement sampling",
abstract = "Deformable image registration poses a highly non-convex optimisation problem. Conventionally, medical image registration techniques rely on continuous optimisation, which is prone to local minima. Recent advances in the mathematics and new programming methods enable these disadvantages to be overcome using discrete optimisation. In this paper, we present a new technique deeds, which employs a discrete dense displacement sampling for the deformable registration of high resolution CT volumes. The image grid is represented as a minimum spanning tree. Given these constraints a global optimum of the cost function can be found efficiently using dynamic programming, which enforces the smoothness of the deformations. Experimental results demonstrate the advantages of deeds: the registration error for the challenging registration of inhale and exhale pulmonary CT scans is significantly lower than for two state-of-the-art registration techniques, especially in the presence of large deformations and sliding motion at lung surfaces.",
author = "Heinrich, {Mattias P.} and Mark Jenkinson and Brady, {Sir Michael} and Schnabel, {Julia A.}",
year = "2012",
month = "1",
day = "1",
language = "English",
isbn = "9783642334535",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "115--122",
editor = "Polina Golland and Nicholas Ayache and Herve Delingette and Kensaku Mori",
booktitle = "Medical Image Computing and Computer-Assisted Intervention, MICCAI2012 - 15th International Conference, Proceedings",

}

Heinrich, MP, Jenkinson, M, Brady, SM & Schnabel, JA 2012, Globally optimal deformable registration on a minimum spanning tree using dense displacement sampling. in P Golland, N Ayache, H Delingette & K Mori (eds), Medical Image Computing and Computer-Assisted Intervention, MICCAI2012 - 15th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7512 LNCS, Springer Verlag, pp. 115-122, 15th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2012, Nice, France, 1/10/12.

Globally optimal deformable registration on a minimum spanning tree using dense displacement sampling. / Heinrich, Mattias P.; Jenkinson, Mark; Brady, Sir Michael; Schnabel, Julia A.

Medical Image Computing and Computer-Assisted Intervention, MICCAI2012 - 15th International Conference, Proceedings. ed. / Polina Golland; Nicholas Ayache; Herve Delingette; Kensaku Mori. Springer Verlag, 2012. p. 115-122 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7512 LNCS).

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

TY - GEN

T1 - Globally optimal deformable registration on a minimum spanning tree using dense displacement sampling

AU - Heinrich, Mattias P.

AU - Jenkinson, Mark

AU - Brady, Sir Michael

AU - Schnabel, Julia A.

PY - 2012/1/1

Y1 - 2012/1/1

N2 - Deformable image registration poses a highly non-convex optimisation problem. Conventionally, medical image registration techniques rely on continuous optimisation, which is prone to local minima. Recent advances in the mathematics and new programming methods enable these disadvantages to be overcome using discrete optimisation. In this paper, we present a new technique deeds, which employs a discrete dense displacement sampling for the deformable registration of high resolution CT volumes. The image grid is represented as a minimum spanning tree. Given these constraints a global optimum of the cost function can be found efficiently using dynamic programming, which enforces the smoothness of the deformations. Experimental results demonstrate the advantages of deeds: the registration error for the challenging registration of inhale and exhale pulmonary CT scans is significantly lower than for two state-of-the-art registration techniques, especially in the presence of large deformations and sliding motion at lung surfaces.

AB - Deformable image registration poses a highly non-convex optimisation problem. Conventionally, medical image registration techniques rely on continuous optimisation, which is prone to local minima. Recent advances in the mathematics and new programming methods enable these disadvantages to be overcome using discrete optimisation. In this paper, we present a new technique deeds, which employs a discrete dense displacement sampling for the deformable registration of high resolution CT volumes. The image grid is represented as a minimum spanning tree. Given these constraints a global optimum of the cost function can be found efficiently using dynamic programming, which enforces the smoothness of the deformations. Experimental results demonstrate the advantages of deeds: the registration error for the challenging registration of inhale and exhale pulmonary CT scans is significantly lower than for two state-of-the-art registration techniques, especially in the presence of large deformations and sliding motion at lung surfaces.

UR - http://www.scopus.com/inward/record.url?scp=84872938499&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9783642334535

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 115

EP - 122

BT - Medical Image Computing and Computer-Assisted Intervention, MICCAI2012 - 15th International Conference, Proceedings

A2 - Golland, Polina

A2 - Ayache, Nicholas

A2 - Delingette, Herve

A2 - Mori, Kensaku

PB - Springer Verlag

ER -

Heinrich MP, Jenkinson M, Brady SM, Schnabel JA. Globally optimal deformable registration on a minimum spanning tree using dense displacement sampling. In Golland P, Ayache N, Delingette H, Mori K, editors, Medical Image Computing and Computer-Assisted Intervention, MICCAI2012 - 15th International Conference, Proceedings. Springer Verlag. 2012. p. 115-122. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).