Segmentation of perivascular spaces using vascular features and structured random forest from 7T MR image

Jun Zhang; Yaozong Gao; Sang Hyun Park; Xiaopeng Zong; Weili Lin; Dinggang Shen

doi:10.1007/978-3-319-47157-0_8

Segmentation of perivascular spaces using vascular features and structured random forest from 7T MR image

Jun Zhang, Yaozong Gao, Sang Hyun Park, Xiaopeng Zong, Weili Lin, Dinggang Shen

Department of Brain and Cognitive Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

10 Citations (Scopus)

Abstract

Quantitative analysis of perivascular spaces (PVSs) is important to reveal the correlations between cerebrovascular lesions and neurodegenerative diseases. In this study, we propose a learning-based segmentation framework to extract the PVSs from high-resolution 7T MR images. Specifically, we integrate three types of vascular filter responses into a structured random forest for classifying voxels into PVS and background. In addition, we also propose a novel entropy-based sampling strategy to extract informative samples in the background for training the classification model. Since various vascular features can be extracted by the three vascular filters, even thin and low-contrast structures can be effectively extracted from the noisy background. Moreover, continuous and smooth segmentation results can be obtained by utilizing the patch-based structured labels. The segmentation performance is evaluated on 19 subjects with 7T MR images, and the experimental results demonstrate that the joint use of entropy-based sampling strategy, vascular features and structured learning improves the segmentation accuracy, with the Dice similarity coefficient reaching 66%.

Original language	English
Title of host publication	Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings
Editors	Li Wang, Heung-Il Suk, Yinghuan Shi, Ehsan Adeli, Qian Wang
Publisher	Springer Verlag
Pages	61-68
Number of pages	8
ISBN (Print)	9783319471563
DOIs	https://doi.org/10.1007/978-3-319-47157-0_8
Publication status	Published - 2016
Event	7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016 - Athens, Greece Duration: 2016 Oct 17 → 2016 Oct 17

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10019 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016
Country/Territory	Greece
City	Athens
Period	16/10/17 → 16/10/17

Bibliographical note

Publisher Copyright:
© Springer International Publishing AG 2016.

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-319-47157-0_8

Cite this

Zhang, J., Gao, Y., Park, S. H., Zong, X., Lin, W., & Shen, D. (2016). Segmentation of perivascular spaces using vascular features and structured random forest from 7T MR image. In L. Wang, H.-I. Suk, Y. Shi, E. Adeli, & Q. Wang (Eds.), Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings (pp. 61-68). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10019 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-47157-0_8

Segmentation of perivascular spaces using vascular features and structured random forest from 7T MR image. / Zhang, Jun; Gao, Yaozong; Park, Sang Hyun et al.
Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings. ed. / Li Wang; Heung-Il Suk; Yinghuan Shi; Ehsan Adeli; Qian Wang. Springer Verlag, 2016. p. 61-68 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10019 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Zhang, J, Gao, Y, Park, SH, Zong, X, Lin, W & Shen, D 2016, Segmentation of perivascular spaces using vascular features and structured random forest from 7T MR image. in L Wang, H-I Suk, Y Shi, E Adeli & Q Wang (eds), Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10019 LNCS, Springer Verlag, pp. 61-68, 7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016, Athens, Greece, 16/10/17. https://doi.org/10.1007/978-3-319-47157-0_8

Zhang J, Gao Y, Park SH, Zong X, Lin W, Shen D. Segmentation of perivascular spaces using vascular features and structured random forest from 7T MR image. In Wang L, Suk HI, Shi Y, Adeli E, Wang Q, editors, Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings. Springer Verlag. 2016. p. 61-68. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-47157-0_8

Zhang, Jun ; Gao, Yaozong ; Park, Sang Hyun et al. / Segmentation of perivascular spaces using vascular features and structured random forest from 7T MR image. Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings. editor / Li Wang ; Heung-Il Suk ; Yinghuan Shi ; Ehsan Adeli ; Qian Wang. Springer Verlag, 2016. pp. 61-68 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{0c6a0afdfa644db08bd0aad7b56a465c,

title = "Segmentation of perivascular spaces using vascular features and structured random forest from 7T MR image",

abstract = "Quantitative analysis of perivascular spaces (PVSs) is important to reveal the correlations between cerebrovascular lesions and neurodegenerative diseases. In this study, we propose a learning-based segmentation framework to extract the PVSs from high-resolution 7T MR images. Specifically, we integrate three types of vascular filter responses into a structured random forest for classifying voxels into PVS and background. In addition, we also propose a novel entropy-based sampling strategy to extract informative samples in the background for training the classification model. Since various vascular features can be extracted by the three vascular filters, even thin and low-contrast structures can be effectively extracted from the noisy background. Moreover, continuous and smooth segmentation results can be obtained by utilizing the patch-based structured labels. The segmentation performance is evaluated on 19 subjects with 7T MR images, and the experimental results demonstrate that the joint use of entropy-based sampling strategy, vascular features and structured learning improves the segmentation accuracy, with the Dice similarity coefficient reaching 66%.",

author = "Jun Zhang and Yaozong Gao and Park, {Sang Hyun} and Xiaopeng Zong and Weili Lin and Dinggang Shen",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2016.; 7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016 ; Conference date: 17-10-2016 Through 17-10-2016",

year = "2016",

doi = "10.1007/978-3-319-47157-0_8",

language = "English",

isbn = "9783319471563",

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

publisher = "Springer Verlag",

pages = "61--68",

editor = "Li Wang and Heung-Il Suk and Yinghuan Shi and Ehsan Adeli and Qian Wang",

booktitle = "Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings",

}

TY - GEN

T1 - Segmentation of perivascular spaces using vascular features and structured random forest from 7T MR image

AU - Zhang, Jun

AU - Gao, Yaozong

AU - Park, Sang Hyun

AU - Zong, Xiaopeng

AU - Lin, Weili

AU - Shen, Dinggang

PY - 2016

Y1 - 2016

N2 - Quantitative analysis of perivascular spaces (PVSs) is important to reveal the correlations between cerebrovascular lesions and neurodegenerative diseases. In this study, we propose a learning-based segmentation framework to extract the PVSs from high-resolution 7T MR images. Specifically, we integrate three types of vascular filter responses into a structured random forest for classifying voxels into PVS and background. In addition, we also propose a novel entropy-based sampling strategy to extract informative samples in the background for training the classification model. Since various vascular features can be extracted by the three vascular filters, even thin and low-contrast structures can be effectively extracted from the noisy background. Moreover, continuous and smooth segmentation results can be obtained by utilizing the patch-based structured labels. The segmentation performance is evaluated on 19 subjects with 7T MR images, and the experimental results demonstrate that the joint use of entropy-based sampling strategy, vascular features and structured learning improves the segmentation accuracy, with the Dice similarity coefficient reaching 66%.

AB - Quantitative analysis of perivascular spaces (PVSs) is important to reveal the correlations between cerebrovascular lesions and neurodegenerative diseases. In this study, we propose a learning-based segmentation framework to extract the PVSs from high-resolution 7T MR images. Specifically, we integrate three types of vascular filter responses into a structured random forest for classifying voxels into PVS and background. In addition, we also propose a novel entropy-based sampling strategy to extract informative samples in the background for training the classification model. Since various vascular features can be extracted by the three vascular filters, even thin and low-contrast structures can be effectively extracted from the noisy background. Moreover, continuous and smooth segmentation results can be obtained by utilizing the patch-based structured labels. The segmentation performance is evaluated on 19 subjects with 7T MR images, and the experimental results demonstrate that the joint use of entropy-based sampling strategy, vascular features and structured learning improves the segmentation accuracy, with the Dice similarity coefficient reaching 66%.

UR - http://www.scopus.com/inward/record.url?scp=84992491329&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-47157-0_8

DO - 10.1007/978-3-319-47157-0_8

M3 - Conference contribution

AN - SCOPUS:84992491329

SN - 9783319471563

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

SP - 61

EP - 68

BT - Machine Learning in Medical Imaging - 7th International Workshop, MLMI 2016 held in conjunction with MICCAI 2016, Proceedings

A2 - Wang, Li

A2 - Suk, Heung-Il

A2 - Shi, Yinghuan

A2 - Adeli, Ehsan

A2 - Wang, Qian

PB - Springer Verlag

T2 - 7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016

Y2 - 17 October 2016 through 17 October 2016

ER -

Segmentation of perivascular spaces using vascular features and structured random forest from 7T MR image

Abstract

Publication series

Other

Bibliographical note

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this