Enhancing Label-Driven Deep Deformable Image Registration with Local Distance Metrics for State-of-the-Art Cardiac Motion Tracking

Alessa Hering; Sven Kuckertz; Stefan Heldmann; Mattias P. Heinrich

doi:10.1007/978-3-658-25326-4_69

Enhancing Label-Driven Deep Deformable Image Registration with Local Distance Metrics for State-of-the-Art Cardiac Motion Tracking

Alessa Hering, Sven Kuckertz, Stefan Heldmann, Mattias P. Heinrich

Abstract

While deep learning has achieved significant advances in accuracy for medical image segmentation, its benefits for deformable image registration have so far remained limited to reduced computation times. Previous work has either focused on replacing the iterative optimization of distance and smoothness terms with CNN-layers or using supervised approaches driven by labels. Our method is the first to combine the complementary strengths of global semantic information (represented by segmentation labels) and local distance metrics that help align surrounding structures. We demonstrate significant higher Dice scores (of 86.5%) for deformable cardiac image registration compared to classic registration (79.0%) as well as label-driven deep learning frameworks (83.4%).

Original language	English
Title of host publication	Bildverarbeitung für die Medizin 2019
Number of pages	6
Publisher	Springer Verlag
Publication date	2019
Pages	309-314
ISBN (Print)	978-3-658-25325-7
ISBN (Electronic)	978-3-658-25326-4
DOIs	https://doi.org/10.1007/978-3-658-25326-4_69
Publication status	Published - 2019
Event	Bildverarbeitung für die Medizin 2019: Algorithmen - Systeme - Anwendungen - Lübeck, Lübeck, Germany Duration: 17.03.2019 → 19.03.2019

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

10.1007/978-3-658-25326-4_69

Cite this

@inproceedings{ba801c13b91248a188c9ff587d32b886,

title = "Enhancing Label-Driven Deep Deformable Image Registration with Local Distance Metrics for State-of-the-Art Cardiac Motion Tracking",

abstract = "While deep learning has achieved significant advances in accuracy for medical image segmentation, its benefits for deformable image registration have so far remained limited to reduced computation times. Previous work has either focused on replacing the iterative optimization of distance and smoothness terms with CNN-layers or using supervised approaches driven by labels. Our method is the first to combine the complementary strengths of global semantic information (represented by segmentation labels) and local distance metrics that help align surrounding structures. We demonstrate significant higher Dice scores (of 86.5\%) for deformable cardiac image registration compared to classic registration (79.0\%) as well as label-driven deep learning frameworks (83.4\%).",

author = "Alessa Hering and Sven Kuckertz and Stefan Heldmann and Heinrich, \{Mattias P.\}",

year = "2019",

doi = "10.1007/978-3-658-25326-4\_69",

language = "English",

isbn = "978-3-658-25325-7",

series = "Informatik aktuell",

publisher = "Springer Verlag",

pages = "309--314",

booktitle = "Bildverarbeitung f{\"u}r die Medizin 2019",

}

Hering, A, Kuckertz, S, Heldmann, S & Heinrich, MP 2019, Enhancing Label-Driven Deep Deformable Image Registration with Local Distance Metrics for State-of-the-Art Cardiac Motion Tracking. in Bildverarbeitung für die Medizin 2019. Informatik aktuell, Springer Verlag, pp. 309-314, Bildverarbeitung für die Medizin 2019, Lübeck, Germany, 17.03.19. https://doi.org/10.1007/978-3-658-25326-4_69

Enhancing Label-Driven Deep Deformable Image Registration with Local Distance Metrics for State-of-the-Art Cardiac Motion Tracking. / Hering, Alessa; Kuckertz, Sven; Heldmann, Stefan et al.
Bildverarbeitung für die Medizin 2019. Springer Verlag, 2019. p. 309-314 (Informatik aktuell).

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

TY - GEN

T1 - Enhancing Label-Driven Deep Deformable Image Registration with Local Distance Metrics for State-of-the-Art Cardiac Motion Tracking

AU - Hering, Alessa

AU - Kuckertz, Sven

AU - Heldmann, Stefan

AU - Heinrich, Mattias P.

PY - 2019

Y1 - 2019

N2 - While deep learning has achieved significant advances in accuracy for medical image segmentation, its benefits for deformable image registration have so far remained limited to reduced computation times. Previous work has either focused on replacing the iterative optimization of distance and smoothness terms with CNN-layers or using supervised approaches driven by labels. Our method is the first to combine the complementary strengths of global semantic information (represented by segmentation labels) and local distance metrics that help align surrounding structures. We demonstrate significant higher Dice scores (of 86.5%) for deformable cardiac image registration compared to classic registration (79.0%) as well as label-driven deep learning frameworks (83.4%).

AB - While deep learning has achieved significant advances in accuracy for medical image segmentation, its benefits for deformable image registration have so far remained limited to reduced computation times. Previous work has either focused on replacing the iterative optimization of distance and smoothness terms with CNN-layers or using supervised approaches driven by labels. Our method is the first to combine the complementary strengths of global semantic information (represented by segmentation labels) and local distance metrics that help align surrounding structures. We demonstrate significant higher Dice scores (of 86.5%) for deformable cardiac image registration compared to classic registration (79.0%) as well as label-driven deep learning frameworks (83.4%).

UR - https://www.scopus.com/pages/publications/85065078034

UR - http://www.mendeley.com/research/enhancing-labeldriven-deep-deformable-image-registration-local-distance-metrics-stateoftheart-cardia

U2 - 10.1007/978-3-658-25326-4_69

DO - 10.1007/978-3-658-25326-4_69

M3 - Conference contribution

SN - 978-3-658-25325-7

T3 - Informatik aktuell

SP - 309

EP - 314

BT - Bildverarbeitung für die Medizin 2019

PB - Springer Verlag

T2 - Bildverarbeitung für die Medizin 2019

Y2 - 17 March 2019 through 19 March 2019

ER -

Enhancing Label-Driven Deep Deformable Image Registration with Local Distance Metrics for State-of-the-Art Cardiac Motion Tracking

Abstract

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this