Memory Efficient LDDMM for Lung CT

Thomas Polzin; Marc Niethammer; Mattias P. Heinrich; Heinz Handels; Jan Modersitzki

doi:10.1007/978-3-319-46726-9_4

Memory Efficient LDDMM for Lung CT

Thomas Polzin^*, Marc Niethammer, Mattias P. Heinrich, Heinz Handels, Jan Modersitzki

^*Corresponding author for this work

University of North Carolina

22 Citations (Scopus)

Abstract

In this paper a novel Large Deformation Diffeomorphic Metric Mapping (LDDMM) scheme is presented which has significantly lower computational and memory demands than standard LDDMM but achieves the same accuracy. We exploit the smoothness of velocities and transformations by using a coarser discretization compared to the image resolution. This reduces required memory and accelerates numerical optimization as well as solution of transport equations. Accuracy is essentially unchanged as the mismatch of transformed moving and fixed image is incorporated into the model at high resolution. Reductions in memory consumption and runtime are demonstrated for registration of lung CT images. State-of-the-art accuracy is shown for the challenging DIRLab chronic obstructive pulmonary disease (COPD) lung CT data sets obtaining a mean landmark distance after registration of 1.03mm and the best average results so far.

Original language	English
Title of host publication	Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings
Editors	Leo Joskowicz, Mert R. Sabuncu, William Wells, Gozde Unal, Sebastian Ourselin
Number of pages	9
Publisher	Springer International Publishing
Publication date	02.10.2016
Pages	28-36
ISBN (Print)	978-3-319-46725-2
ISBN (Electronic)	978-3-319-46726-9
DOIs	https://doi.org/10.1007/978-3-319-46726-9_4 https://doi.org/10.1007/978-3-319-46726-9_4
Publication status	Published - 02.10.2016
Event	19th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2016 - Athens, Greece Duration: 17.10.2016 → 21.10.2016 http://miccai2016.org/en/

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being
SDG 9 Industry, Innovation, and Infrastructure

Access to Document

Cite this

Polzin, T., Niethammer, M., Heinrich, M. P., Handels, H., & Modersitzki, J. (2016). Memory Efficient LDDMM for Lung CT. In L. Joskowicz, M. R. Sabuncu, W. Wells, G. Unal, & S. Ourselin (Eds.), Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings (pp. 28-36). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9902 LNCS). Springer International Publishing. https://doi.org/10.1007/978-3-319-46726-9_4, https://doi.org/10.1007/978-3-319-46726-9_4

Polzin, Thomas ; Niethammer, Marc ; Heinrich, Mattias P. et al. / Memory Efficient LDDMM for Lung CT. Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. editor / Leo Joskowicz ; Mert R. Sabuncu ; William Wells ; Gozde Unal ; Sebastian Ourselin. Springer International Publishing, 2016. pp. 28-36 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{ef93917b40d24d7a93ae6803928981bd,

title = "Memory Efficient LDDMM for Lung CT",

abstract = "In this paper a novel Large Deformation Diffeomorphic Metric Mapping (LDDMM) scheme is presented which has significantly lower computational and memory demands than standard LDDMM but achieves the same accuracy. We exploit the smoothness of velocities and transformations by using a coarser discretization compared to the image resolution. This reduces required memory and accelerates numerical optimization as well as solution of transport equations. Accuracy is essentially unchanged as the mismatch of transformed moving and fixed image is incorporated into the model at high resolution. Reductions in memory consumption and runtime are demonstrated for registration of lung CT images. State-of-the-art accuracy is shown for the challenging DIRLab chronic obstructive pulmonary disease (COPD) lung CT data sets obtaining a mean landmark distance after registration of 1.03mm and the best average results so far.",

author = "Thomas Polzin and Marc Niethammer and Heinrich, \{Mattias P.\} and Heinz Handels and Jan Modersitzki",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2016.; 19th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2016<br/> ; Conference date: 17-10-2016 Through 21-10-2016",

year = "2016",

month = oct,

day = "2",

doi = "10.1007/978-3-319-46726-9\_4",

language = "English",

isbn = "978-3-319-46725-2",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer International Publishing",

pages = "28--36",

editor = "Leo Joskowicz and Sabuncu, \{Mert R.\} and William Wells and Gozde Unal and Sebastian Ourselin",

booktitle = "Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings",

url = "http://miccai2016.org/en/",

}

Polzin, T, Niethammer, M, Heinrich, MP , Handels, H & Modersitzki, J 2016, Memory Efficient LDDMM for Lung CT. in L Joskowicz, MR Sabuncu, W Wells, G Unal & S Ourselin (eds), Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9902 LNCS, Springer International Publishing, pp. 28-36, 19th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2016
, Athens, Greece, 17.10.16. https://doi.org/10.1007/978-3-319-46726-9_4, https://doi.org/10.1007/978-3-319-46726-9_4

Memory Efficient LDDMM for Lung CT. / Polzin, Thomas; Niethammer, Marc; Heinrich, Mattias P. et al.
Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. ed. / Leo Joskowicz; Mert R. Sabuncu; William Wells; Gozde Unal; Sebastian Ourselin. Springer International Publishing, 2016. p. 28-36 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9902 LNCS).

Research output: Chapters in Books/Reports/Conference Proceedings › Conference contribution › peer-review

TY - GEN

T1 - Memory Efficient LDDMM for Lung CT

AU - Polzin, Thomas

AU - Niethammer, Marc

AU - Heinrich, Mattias P.

AU - Handels, Heinz

AU - Modersitzki, Jan

N1 - Publisher Copyright: © Springer International Publishing AG 2016.

PY - 2016/10/2

Y1 - 2016/10/2

N2 - In this paper a novel Large Deformation Diffeomorphic Metric Mapping (LDDMM) scheme is presented which has significantly lower computational and memory demands than standard LDDMM but achieves the same accuracy. We exploit the smoothness of velocities and transformations by using a coarser discretization compared to the image resolution. This reduces required memory and accelerates numerical optimization as well as solution of transport equations. Accuracy is essentially unchanged as the mismatch of transformed moving and fixed image is incorporated into the model at high resolution. Reductions in memory consumption and runtime are demonstrated for registration of lung CT images. State-of-the-art accuracy is shown for the challenging DIRLab chronic obstructive pulmonary disease (COPD) lung CT data sets obtaining a mean landmark distance after registration of 1.03mm and the best average results so far.

AB - In this paper a novel Large Deformation Diffeomorphic Metric Mapping (LDDMM) scheme is presented which has significantly lower computational and memory demands than standard LDDMM but achieves the same accuracy. We exploit the smoothness of velocities and transformations by using a coarser discretization compared to the image resolution. This reduces required memory and accelerates numerical optimization as well as solution of transport equations. Accuracy is essentially unchanged as the mismatch of transformed moving and fixed image is incorporated into the model at high resolution. Reductions in memory consumption and runtime are demonstrated for registration of lung CT images. State-of-the-art accuracy is shown for the challenging DIRLab chronic obstructive pulmonary disease (COPD) lung CT data sets obtaining a mean landmark distance after registration of 1.03mm and the best average results so far.

UR - http://www.imi.uni-luebeck.de/en/content/memory-efficient-lddmm-lung-ct

U2 - 10.1007/978-3-319-46726-9_4

DO - 10.1007/978-3-319-46726-9_4

M3 - Conference contribution

SN - 978-3-319-46725-2

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

SP - 28

EP - 36

BT - Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings

A2 - Joskowicz, Leo

A2 - Sabuncu, Mert R.

A2 - Wells, William

A2 - Unal, Gozde

A2 - Ourselin, Sebastian

PB - Springer International Publishing

T2 - 19th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2016<br/>

Y2 - 17 October 2016 through 21 October 2016

ER -

Polzin T, Niethammer M, Heinrich MP , Handels H , Modersitzki J. Memory Efficient LDDMM for Lung CT. In Joskowicz L, Sabuncu MR, Wells W, Unal G, Ourselin S, editors, Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. Springer International Publishing. 2016. p. 28-36. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-46726-9_4, 10.1007/978-3-319-46726-9_4

Memory Efficient LDDMM for Lung CT

Abstract

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this