Automatic intervertebral disc localization and segmentation in 3D MR images based on regression forests and active contours

Martin Urschler; Kerstin Hammernik; Thomas Ebner; Darko Stern

doi:10.1007/978-3-319-41827-8_13

Automatic intervertebral disc localization and segmentation in 3D MR images based on regression forests and active contours

Martin Urschler^*, Kerstin Hammernik, Thomas Ebner, Darko Stern

^*Korrespondierende/r Autor/-in für diese Arbeit

Institut für Maschinelles Sehen und Darstellen (7100)

Publikation: Beitrag in Buch/Bericht/Konferenzband › Beitrag in Buch/Bericht › Begutachtung

Abstract

We introduce a fully automatic localization and segmentation pipeline for three-dimensional (3D) intervertebral discs (IVDs), consisting of a regression-based prediction of vertebral bodies and IVD positions as well as a 3D geodesic active contour segmentation delineating the IVDs. The approach was evaluated on the data set of the challenge in conjunction with the 3^rd MICCAI Workshop & Challenge on Computational Methods and Clinical Applications for Spine Imaging -MICCAI– CSI2015, that consists of 15 magnetic resonance images of the lumbar spine with given ground truth segmentations. Based on a localization accuracy of 3.9±1.6 mm, we achieve segmentation results in terms of the Dice similarity coefficient of 89.1 ±2.9% averaged over the whole data set.

Originalsprache	englisch
Titel	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Herausgeber (Verlag)	Springer Verlag Heidelberg
Seiten	130-140
Seitenumfang	11
Band	9402
ISBN (Print)	9783319418261
DOIs	https://doi.org/10.1007/978-3-319-41827-8_13
Publikationsstatus	Veröffentlicht - 2016
Veranstaltung	International Conference on Medical Image Computing and Computer Assisted Intervention: MICCAI 2015 - Munich, Deutschland Dauer: 5 Okt. 2015 → 9 Okt. 2015

Publikationsreihe

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Band	9402
ISSN (Print)	03029743
ISSN (elektronisch)	16113349

Konferenz

Konferenz	International Conference on Medical Image Computing and Computer Assisted Intervention
Land/Gebiet	Deutschland
Ort	Munich
Zeitraum	5/10/15 → 9/10/15

ASJC Scopus subject areas

Allgemeine Computerwissenschaft
Theoretische Informatik

Fields of Expertise

Information, Communication & Computing

Kooperationen

BioTechMed-Graz

Zugriff auf Dokument

10.1007/978-3-319-41827-8_13

Andere Dateien und Links

http://www.scopus.com/inward/record.url?scp=84979211566&partnerID=8YFLogxK

FWF - FAME - Vollautomatische MRT-basierte Altersschätzung von Jugendlichen
Bischof, H. & Urschler, M.
1/07/15 → 31/12/18
Projekt: Forschungsprojekt

Dieses zitieren

Urschler, M., Hammernik, K., Ebner, T., & Stern, D. (2016). Automatic intervertebral disc localization and segmentation in 3D MR images based on regression forests and active contours. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Band 9402, S. 130-140). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 9402). Springer Verlag Heidelberg. https://doi.org/10.1007/978-3-319-41827-8_13

Automatic intervertebral disc localization and segmentation in 3D MR images based on regression forests and active contours. / Urschler, Martin; Hammernik, Kerstin; Ebner, Thomas et al.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Band 9402 Springer Verlag Heidelberg, 2016. S. 130-140 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 9402).

Publikation: Beitrag in Buch/Bericht/Konferenzband › Beitrag in Buch/Bericht › Begutachtung

Urschler, M, Hammernik, K, Ebner, T & Stern, D 2016, Automatic intervertebral disc localization and segmentation in 3D MR images based on regression forests and active contours. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Bd. 9402, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bd. 9402, Springer Verlag Heidelberg, S. 130-140, International Conference on Medical Image Computing and Computer Assisted Intervention, Munich, Deutschland, 5/10/15. https://doi.org/10.1007/978-3-319-41827-8_13

Urschler M, Hammernik K, Ebner T, Stern D. Automatic intervertebral disc localization and segmentation in 3D MR images based on regression forests and active contours. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Band 9402. Springer Verlag Heidelberg. 2016. S. 130-140. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-41827-8_13

Urschler, Martin ; Hammernik, Kerstin ; Ebner, Thomas et al. / Automatic intervertebral disc localization and segmentation in 3D MR images based on regression forests and active contours. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Band 9402 Springer Verlag Heidelberg, 2016. S. 130-140 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inbook{74dde05e1b4b446b89ad5e94161cbd1e,

title = "Automatic intervertebral disc localization and segmentation in 3D MR images based on regression forests and active contours",

abstract = "We introduce a fully automatic localization and segmentation pipeline for three-dimensional (3D) intervertebral discs (IVDs), consisting of a regression-based prediction of vertebral bodies and IVD positions as well as a 3D geodesic active contour segmentation delineating the IVDs. The approach was evaluated on the data set of the challenge in conjunction with the 3rd MICCAI Workshop & Challenge on Computational Methods and Clinical Applications for Spine Imaging -MICCAI– CSI2015, that consists of 15 magnetic resonance images of the lumbar spine with given ground truth segmentations. Based on a localization accuracy of 3.9±1.6 mm, we achieve segmentation results in terms of the Dice similarity coefficient of 89.1 ±2.9% averaged over the whole data set.",

author = "Martin Urschler and Kerstin Hammernik and Thomas Ebner and Darko Stern",

year = "2016",

doi = "10.1007/978-3-319-41827-8_13",

language = "English",

isbn = "9783319418261",

volume = "9402",

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

publisher = "Springer Verlag Heidelberg",

pages = "130--140",

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

note = "International Conference on Medical Image Computing and Computer Assisted Intervention : MICCAI 2015 ; Conference date: 05-10-2015 Through 09-10-2015",

}

TY - CHAP

T1 - Automatic intervertebral disc localization and segmentation in 3D MR images based on regression forests and active contours

AU - Urschler, Martin

AU - Hammernik, Kerstin

AU - Ebner, Thomas

AU - Stern, Darko

PY - 2016

Y1 - 2016

N2 - We introduce a fully automatic localization and segmentation pipeline for three-dimensional (3D) intervertebral discs (IVDs), consisting of a regression-based prediction of vertebral bodies and IVD positions as well as a 3D geodesic active contour segmentation delineating the IVDs. The approach was evaluated on the data set of the challenge in conjunction with the 3rd MICCAI Workshop & Challenge on Computational Methods and Clinical Applications for Spine Imaging -MICCAI– CSI2015, that consists of 15 magnetic resonance images of the lumbar spine with given ground truth segmentations. Based on a localization accuracy of 3.9±1.6 mm, we achieve segmentation results in terms of the Dice similarity coefficient of 89.1 ±2.9% averaged over the whole data set.

AB - We introduce a fully automatic localization and segmentation pipeline for three-dimensional (3D) intervertebral discs (IVDs), consisting of a regression-based prediction of vertebral bodies and IVD positions as well as a 3D geodesic active contour segmentation delineating the IVDs. The approach was evaluated on the data set of the challenge in conjunction with the 3rd MICCAI Workshop & Challenge on Computational Methods and Clinical Applications for Spine Imaging -MICCAI– CSI2015, that consists of 15 magnetic resonance images of the lumbar spine with given ground truth segmentations. Based on a localization accuracy of 3.9±1.6 mm, we achieve segmentation results in terms of the Dice similarity coefficient of 89.1 ±2.9% averaged over the whole data set.

UR - http://www.scopus.com/inward/record.url?scp=84979211566&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-41827-8_13

DO - 10.1007/978-3-319-41827-8_13

M3 - Chapter

AN - SCOPUS:84979211566

SN - 9783319418261

VL - 9402

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

SP - 130

EP - 140

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

PB - Springer Verlag Heidelberg

T2 - International Conference on Medical Image Computing and Computer Assisted Intervention

Y2 - 5 October 2015 through 9 October 2015

ER -

Automatic intervertebral disc localization and segmentation in 3D MR images based on regression forests and active contours

Abstract

Publikationsreihe

Konferenz

ASJC Scopus subject areas

Fields of Expertise

Kooperationen

Zugriff auf Dokument

Andere Dateien und Links

Fingerprint

Projekte

FWF - FAME - Vollautomatische MRT-basierte Altersschätzung von Jugendlichen

Dieses zitieren