Automatic segmentation of vertebrae from radiographs: a sample-driven active shape model approach

Peter Mysling; Peter Kersten Petersen; Mads Nielsen; Martin Lillholm

doi:10.1007/978-3-642-24319-6_2

Automatic segmentation of vertebrae from radiographs: a sample-driven active shape model approach

Peter Mysling, Peter Kersten Petersen, Mads Nielsen, Martin Lillholm

Department of Computer Science

3 Citations (Scopus)

Abstract

Segmentation of vertebral contours is an essential task in the design of automatic tools for vertebral fracture assessment. In this paper, we propose a novel segmentation technique which does not require operator interaction. The proposed technique solves the segmentation problem in a hierarchical manner. In a first phase, a coarse estimate of the overall spine alignment and the vertebra locations is computed using a shape model sampling scheme. These samples are used to initialize a second phase of active shape model search, under a nonlinear model of vertebra appearance. The search is constrained by a conditional shape model, based on the variability of the coarse spine location estimates. The technique is evaluated on a data set of manually annotated lumbar radiographs. The results compare favorably to the previous work in automatic vertebra segmentation, in terms of both segmentation accuracy and failure rate.

Original language	English
Title of host publication	Machine Learning in Medical Imaging : Second International Workshop, MLMI 2011, Held in Conjunction with MICCAI 2011, Toronto, Canada, September 18, 2011. Proceedings
Editors	Kenji Suzuki, Fei Wang, Dinggang Shen, Pingkun Yan
Number of pages	8
Publisher	Springer
Publication date	2011
Pages	10-17
ISBN (Print)	978-3-642-24318-9
ISBN (Electronic)	978-3-642-24319-6
DOIs	https://doi.org/10.1007/978-3-642-24319-6_2
Publication status	Published - 2011
Event	International Workshop on Machine Learning in Medical Imaging - Toronto, Canada Duration: 18 Sept 2011 → 18 Sept 2011 Conference number: 2

Conference

Conference	International Workshop on Machine Learning in Medical Imaging
Number	2
Country/Territory	Canada
City	Toronto
Period	18/09/2011 → 18/09/2011

Series	Lecture notes in computer science
Volume	7009
ISSN	0302-9743

Access to Document

10.1007/978-3-642-24319-6_2

Cite this

Mysling, P., Petersen, P. K., Nielsen, M., & Lillholm, M. (2011). Automatic segmentation of vertebrae from radiographs: a sample-driven active shape model approach. In K. Suzuki, F. Wang, D. Shen, & P. Yan (Eds.), Machine Learning in Medical Imaging: Second International Workshop, MLMI 2011, Held in Conjunction with MICCAI 2011, Toronto, Canada, September 18, 2011. Proceedings (pp. 10-17). Springer. https://doi.org/10.1007/978-3-642-24319-6_2

Automatic segmentation of vertebrae from radiographs: a sample-driven active shape model approach. / Mysling, Peter; Petersen, Peter Kersten; Nielsen, Mads et al.
Machine Learning in Medical Imaging: Second International Workshop, MLMI 2011, Held in Conjunction with MICCAI 2011, Toronto, Canada, September 18, 2011. Proceedings. ed. / Kenji Suzuki; Fei Wang; Dinggang Shen; Pingkun Yan. Springer, 2011. p. 10-17 (Lecture notes in computer science, Vol. 7009).

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Mysling, P, Petersen, PK, Nielsen, M & Lillholm, M 2011, Automatic segmentation of vertebrae from radiographs: a sample-driven active shape model approach. in K Suzuki, F Wang, D Shen & P Yan (eds), Machine Learning in Medical Imaging: Second International Workshop, MLMI 2011, Held in Conjunction with MICCAI 2011, Toronto, Canada, September 18, 2011. Proceedings. Springer, Lecture notes in computer science, vol. 7009, pp. 10-17, International Workshop on Machine Learning in Medical Imaging, Toronto, Canada, 18/09/2011. https://doi.org/10.1007/978-3-642-24319-6_2

Mysling P, Petersen PK, Nielsen M , Lillholm M. Automatic segmentation of vertebrae from radiographs: a sample-driven active shape model approach. In Suzuki K, Wang F, Shen D, Yan P, editors, Machine Learning in Medical Imaging: Second International Workshop, MLMI 2011, Held in Conjunction with MICCAI 2011, Toronto, Canada, September 18, 2011. Proceedings. Springer. 2011. p. 10-17. (Lecture notes in computer science, Vol. 7009). doi: 10.1007/978-3-642-24319-6_2

Mysling, Peter ; Petersen, Peter Kersten ; Nielsen, Mads et al. / Automatic segmentation of vertebrae from radiographs : a sample-driven active shape model approach. Machine Learning in Medical Imaging: Second International Workshop, MLMI 2011, Held in Conjunction with MICCAI 2011, Toronto, Canada, September 18, 2011. Proceedings. editor / Kenji Suzuki ; Fei Wang ; Dinggang Shen ; Pingkun Yan. Springer, 2011. pp. 10-17 (Lecture notes in computer science, Vol. 7009).

@inproceedings{62919a845a894332bd3557e5ce038186,

title = "Automatic segmentation of vertebrae from radiographs: a sample-driven active shape model approach",

abstract = "Segmentation of vertebral contours is an essential task in the design of automatic tools for vertebral fracture assessment. In this paper, we propose a novel segmentation technique which does not require operator interaction. The proposed technique solves the segmentation problem in a hierarchical manner. In a first phase, a coarse estimate of the overall spine alignment and the vertebra locations is computed using a shape model sampling scheme. These samples are used to initialize a second phase of active shape model search, under a nonlinear model of vertebra appearance. The search is constrained by a conditional shape model, based on the variability of the coarse spine location estimates. The technique is evaluated on a data set of manually annotated lumbar radiographs. The results compare favorably to the previous work in automatic vertebra segmentation, in terms of both segmentation accuracy and failure rate. ",

author = "Peter Mysling and Petersen, {Peter Kersten} and Mads Nielsen and Martin Lillholm",

note = "Del af 14th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI); International Workshop on Machine Learning in Medical Imaging, MLMI ; Conference date: 18-09-2011 Through 18-09-2011",

year = "2011",

doi = "10.1007/978-3-642-24319-6_2",

language = "English",

isbn = "978-3-642-24318-9",

series = "Lecture notes in computer science",

publisher = "Springer",

pages = "10--17",

editor = "Kenji Suzuki and Fei Wang and Dinggang Shen and Pingkun Yan",

booktitle = "Machine Learning in Medical Imaging",

}

TY - GEN

T1 - Automatic segmentation of vertebrae from radiographs

T2 - International Workshop on Machine Learning in Medical Imaging

AU - Mysling, Peter

AU - Petersen, Peter Kersten

AU - Nielsen, Mads

AU - Lillholm, Martin

N1 - Conference code: 2

PY - 2011

Y1 - 2011

N2 - Segmentation of vertebral contours is an essential task in the design of automatic tools for vertebral fracture assessment. In this paper, we propose a novel segmentation technique which does not require operator interaction. The proposed technique solves the segmentation problem in a hierarchical manner. In a first phase, a coarse estimate of the overall spine alignment and the vertebra locations is computed using a shape model sampling scheme. These samples are used to initialize a second phase of active shape model search, under a nonlinear model of vertebra appearance. The search is constrained by a conditional shape model, based on the variability of the coarse spine location estimates. The technique is evaluated on a data set of manually annotated lumbar radiographs. The results compare favorably to the previous work in automatic vertebra segmentation, in terms of both segmentation accuracy and failure rate.

AB - Segmentation of vertebral contours is an essential task in the design of automatic tools for vertebral fracture assessment. In this paper, we propose a novel segmentation technique which does not require operator interaction. The proposed technique solves the segmentation problem in a hierarchical manner. In a first phase, a coarse estimate of the overall spine alignment and the vertebra locations is computed using a shape model sampling scheme. These samples are used to initialize a second phase of active shape model search, under a nonlinear model of vertebra appearance. The search is constrained by a conditional shape model, based on the variability of the coarse spine location estimates. The technique is evaluated on a data set of manually annotated lumbar radiographs. The results compare favorably to the previous work in automatic vertebra segmentation, in terms of both segmentation accuracy and failure rate.

U2 - 10.1007/978-3-642-24319-6_2

DO - 10.1007/978-3-642-24319-6_2

M3 - Article in proceedings

SN - 978-3-642-24318-9

T3 - Lecture notes in computer science

SP - 10

EP - 17

BT - Machine Learning in Medical Imaging

A2 - Suzuki, Kenji

A2 - Wang, Fei

A2 - Shen, Dinggang

A2 - Yan, Pingkun

PB - Springer

Y2 - 18 September 2011 through 18 September 2011

ER -

Automatic segmentation of vertebrae from radiographs: a sample-driven active shape model approach

Abstract

Conference

Access to Document

Fingerprint

Cite this