Prostate segmentation in MR images using ensemble deep convolutional neural networks

Haozhe Jia; Yong Xia; Weidong Cai; Michael Fulham; David Dagan Feng

doi:10.1109/ISBI.2017.7950630

Prostate segmentation in MR images using ensemble deep convolutional neural networks

Haozhe Jia, Yong Xia, Weidong Cai, Michael Fulham, David Dagan Feng

School of Computer Science

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

25 Scopus citations

Abstract

The automated segmentation of the prostate gland from MR images is increasingly used for clinical diagnosis. Since deep learning demonstrates superior performance in computer vision applications, we propose a coarse-to-fine segmentation strategy using ensemble deep convolutional neural networks (DCNNs) to address prostate segmentation in MR images. First, we use registration-based coarse segmentation on pre-processed prostate MR images to define the potential boundary region. We then train four DCNNs as voxel-based classifiers and classify the voxel in the potential region is a prostate voxel when at least three DCNNs made that decision. Finally, we use boundary refinement to eliminate the outliers and smooth the boundary. We evaluated our approach on the MICCAI PROMIS12 challenge dataset and our experimental results verify the effectiveness of the proposed algorithms.

Original language	English
Title of host publication	2017 IEEE 14th International Symposium on Biomedical Imaging, ISBI 2017
Publisher	IEEE Computer Society
Pages	762-765
Number of pages	4
ISBN (Electronic)	9781509011711
DOIs	https://doi.org/10.1109/ISBI.2017.7950630
State	Published - 15 Jun 2017
Event	14th IEEE International Symposium on Biomedical Imaging, ISBI 2017 - Melbourne, Australia Duration: 18 Apr 2017 → 21 Apr 2017

Publication series

Name	Proceedings - International Symposium on Biomedical Imaging
ISSN (Print)	1945-7928
ISSN (Electronic)	1945-8452

Conference

Conference	14th IEEE International Symposium on Biomedical Imaging, ISBI 2017
Country/Territory	Australia
City	Melbourne
Period	18/04/17 → 21/04/17

Keywords

Deep convolutional neural network
MR prostate segmentation
Voxel classification

Access to Document

10.1109/ISBI.2017.7950630

Cite this

Jia, H., Xia, Y., Cai, W., Fulham, M., & Feng, D. D. (2017). Prostate segmentation in MR images using ensemble deep convolutional neural networks. In 2017 IEEE 14th International Symposium on Biomedical Imaging, ISBI 2017 (pp. 762-765). Article 7950630 (Proceedings - International Symposium on Biomedical Imaging). IEEE Computer Society. https://doi.org/10.1109/ISBI.2017.7950630

@inproceedings{c76e7aeac7d5491d9032f9bbf62baee2,

title = "Prostate segmentation in MR images using ensemble deep convolutional neural networks",

abstract = "The automated segmentation of the prostate gland from MR images is increasingly used for clinical diagnosis. Since deep learning demonstrates superior performance in computer vision applications, we propose a coarse-to-fine segmentation strategy using ensemble deep convolutional neural networks (DCNNs) to address prostate segmentation in MR images. First, we use registration-based coarse segmentation on pre-processed prostate MR images to define the potential boundary region. We then train four DCNNs as voxel-based classifiers and classify the voxel in the potential region is a prostate voxel when at least three DCNNs made that decision. Finally, we use boundary refinement to eliminate the outliers and smooth the boundary. We evaluated our approach on the MICCAI PROMIS12 challenge dataset and our experimental results verify the effectiveness of the proposed algorithms.",

keywords = "Deep convolutional neural network, MR prostate segmentation, Voxel classification",

author = "Haozhe Jia and Yong Xia and Weidong Cai and Michael Fulham and Feng, {David Dagan}",

note = "Publisher Copyright: {\textcopyright} 2017 IEEE.; 14th IEEE International Symposium on Biomedical Imaging, ISBI 2017 ; Conference date: 18-04-2017 Through 21-04-2017",

year = "2017",

month = jun,

day = "15",

doi = "10.1109/ISBI.2017.7950630",

language = "英语",

series = "Proceedings - International Symposium on Biomedical Imaging",

publisher = "IEEE Computer Society",

pages = "762--765",

booktitle = "2017 IEEE 14th International Symposium on Biomedical Imaging, ISBI 2017",

}

Jia, H, Xia, Y, Cai, W, Fulham, M & Feng, DD 2017, Prostate segmentation in MR images using ensemble deep convolutional neural networks. in 2017 IEEE 14th International Symposium on Biomedical Imaging, ISBI 2017., 7950630, Proceedings - International Symposium on Biomedical Imaging, IEEE Computer Society, pp. 762-765, 14th IEEE International Symposium on Biomedical Imaging, ISBI 2017, Melbourne, Australia, 18/04/17. https://doi.org/10.1109/ISBI.2017.7950630

Prostate segmentation in MR images using ensemble deep convolutional neural networks. / Jia, Haozhe; Xia, Yong; Cai, Weidong et al.
2017 IEEE 14th International Symposium on Biomedical Imaging, ISBI 2017. IEEE Computer Society, 2017. p. 762-765 7950630 (Proceedings - International Symposium on Biomedical Imaging).

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

TY - GEN

T1 - Prostate segmentation in MR images using ensemble deep convolutional neural networks

AU - Jia, Haozhe

AU - Xia, Yong

AU - Cai, Weidong

AU - Fulham, Michael

AU - Feng, David Dagan

PY - 2017/6/15

Y1 - 2017/6/15

N2 - The automated segmentation of the prostate gland from MR images is increasingly used for clinical diagnosis. Since deep learning demonstrates superior performance in computer vision applications, we propose a coarse-to-fine segmentation strategy using ensemble deep convolutional neural networks (DCNNs) to address prostate segmentation in MR images. First, we use registration-based coarse segmentation on pre-processed prostate MR images to define the potential boundary region. We then train four DCNNs as voxel-based classifiers and classify the voxel in the potential region is a prostate voxel when at least three DCNNs made that decision. Finally, we use boundary refinement to eliminate the outliers and smooth the boundary. We evaluated our approach on the MICCAI PROMIS12 challenge dataset and our experimental results verify the effectiveness of the proposed algorithms.

AB - The automated segmentation of the prostate gland from MR images is increasingly used for clinical diagnosis. Since deep learning demonstrates superior performance in computer vision applications, we propose a coarse-to-fine segmentation strategy using ensemble deep convolutional neural networks (DCNNs) to address prostate segmentation in MR images. First, we use registration-based coarse segmentation on pre-processed prostate MR images to define the potential boundary region. We then train four DCNNs as voxel-based classifiers and classify the voxel in the potential region is a prostate voxel when at least three DCNNs made that decision. Finally, we use boundary refinement to eliminate the outliers and smooth the boundary. We evaluated our approach on the MICCAI PROMIS12 challenge dataset and our experimental results verify the effectiveness of the proposed algorithms.

KW - Deep convolutional neural network

KW - MR prostate segmentation

KW - Voxel classification

UR - http://www.scopus.com/inward/record.url?scp=85023165512&partnerID=8YFLogxK

U2 - 10.1109/ISBI.2017.7950630

DO - 10.1109/ISBI.2017.7950630

M3 - 会议稿件

AN - SCOPUS:85023165512

T3 - Proceedings - International Symposium on Biomedical Imaging

SP - 762

EP - 765

BT - 2017 IEEE 14th International Symposium on Biomedical Imaging, ISBI 2017

PB - IEEE Computer Society

T2 - 14th IEEE International Symposium on Biomedical Imaging, ISBI 2017

Y2 - 18 April 2017 through 21 April 2017

ER -

Prostate segmentation in MR images using ensemble deep convolutional neural networks

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this