An effective CNN approach for vertebrae segmentation from 3D CT images

Chan Pang Kuok, Jin Yuan Hsue, Ting Li Shen, Bing Feng Huang, Chi Yeh Chen, Yung Nien Sun

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Spine surgery is risky when treating spinal diseases or injuries. In order to decrease the damage of tissue and accelerate the recovery, minimally invasive surgery is strongly desired. Because of this, the preoperative planning and intraoperative guidance become very important. A 3D vertebrae model of the surgical site will be built for these purposes from the computer tomography (CT) images. The segmentation methods of vertebrae from the CT images for modeling are widely concerned. The conventional methods were satisfactory but they usually required prior knowledge of the vertebrae. Recently, the deep learning convolutional neural network (CNN) showed an outstanding performance on the end-to-end semantic image segmentation. Benefit from this, we propose an effective and simple convolution neural network based approach for vertebrae segmentation. Different from training and testing on large 3D data, the axial view images are used effectively in this study. A sequence of simple image processing steps are applied for the intervertebral plane detection. Five CT cases are used on this study and the segmentation results are highly accurate with average dice similarity coefficient over 0.95.

Original languageEnglish
Title of host publicationProceedings of the 2018 Pacific Neighborhood Consortium Annual Conference and Joint Meetings
Subtitle of host publicationHuman Rights in Cyberspace, PNC 2018
EditorsShih-Lung Shaw, Ta-Chien Chan, Ling-Jyh Chen
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages7-12
Number of pages6
ISBN (Electronic)9789869531719
DOIs
Publication statusPublished - 2018 Dec 14
Event2018 Pacific Neighborhood Consortium Annual Conference and Joint Meetings, PNC 2018 - San Francisco, United States
Duration: 2018 Oct 272018 Oct 30

Other

Other2018 Pacific Neighborhood Consortium Annual Conference and Joint Meetings, PNC 2018
CountryUnited States
CitySan Francisco
Period18-10-2718-10-30

Fingerprint

neural network
Tomography
Neural networks
Surgery
surgery
Convolution
Image segmentation
Image processing
Semantics
Tissue
Planning
Recovery
damages
semantics
Testing
Disease
planning
segmentation
Segmentation
knowledge

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Information Systems and Management
  • Archaeology
  • Computer Networks and Communications
  • Sociology and Political Science
  • Public Administration

Cite this

Kuok, C. P., Hsue, J. Y., Shen, T. L., Huang, B. F., Chen, C. Y., & Sun, Y. N. (2018). An effective CNN approach for vertebrae segmentation from 3D CT images. In S-L. Shaw, T-C. Chan, & L-J. Chen (Eds.), Proceedings of the 2018 Pacific Neighborhood Consortium Annual Conference and Joint Meetings: Human Rights in Cyberspace, PNC 2018 (pp. 7-12). [8579455] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.23919/PNC.2018.8579455
Kuok, Chan Pang ; Hsue, Jin Yuan ; Shen, Ting Li ; Huang, Bing Feng ; Chen, Chi Yeh ; Sun, Yung Nien. / An effective CNN approach for vertebrae segmentation from 3D CT images. Proceedings of the 2018 Pacific Neighborhood Consortium Annual Conference and Joint Meetings: Human Rights in Cyberspace, PNC 2018. editor / Shih-Lung Shaw ; Ta-Chien Chan ; Ling-Jyh Chen. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 7-12
@inproceedings{17c089719a4a4d1eadf45e6410b3053c,
title = "An effective CNN approach for vertebrae segmentation from 3D CT images",
abstract = "Spine surgery is risky when treating spinal diseases or injuries. In order to decrease the damage of tissue and accelerate the recovery, minimally invasive surgery is strongly desired. Because of this, the preoperative planning and intraoperative guidance become very important. A 3D vertebrae model of the surgical site will be built for these purposes from the computer tomography (CT) images. The segmentation methods of vertebrae from the CT images for modeling are widely concerned. The conventional methods were satisfactory but they usually required prior knowledge of the vertebrae. Recently, the deep learning convolutional neural network (CNN) showed an outstanding performance on the end-to-end semantic image segmentation. Benefit from this, we propose an effective and simple convolution neural network based approach for vertebrae segmentation. Different from training and testing on large 3D data, the axial view images are used effectively in this study. A sequence of simple image processing steps are applied for the intervertebral plane detection. Five CT cases are used on this study and the segmentation results are highly accurate with average dice similarity coefficient over 0.95.",
author = "Kuok, {Chan Pang} and Hsue, {Jin Yuan} and Shen, {Ting Li} and Huang, {Bing Feng} and Chen, {Chi Yeh} and Sun, {Yung Nien}",
year = "2018",
month = "12",
day = "14",
doi = "10.23919/PNC.2018.8579455",
language = "English",
pages = "7--12",
editor = "Shih-Lung Shaw and Ta-Chien Chan and Ling-Jyh Chen",
booktitle = "Proceedings of the 2018 Pacific Neighborhood Consortium Annual Conference and Joint Meetings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

Kuok, CP, Hsue, JY, Shen, TL, Huang, BF, Chen, CY & Sun, YN 2018, An effective CNN approach for vertebrae segmentation from 3D CT images. in S-L Shaw, T-C Chan & L-J Chen (eds), Proceedings of the 2018 Pacific Neighborhood Consortium Annual Conference and Joint Meetings: Human Rights in Cyberspace, PNC 2018., 8579455, Institute of Electrical and Electronics Engineers Inc., pp. 7-12, 2018 Pacific Neighborhood Consortium Annual Conference and Joint Meetings, PNC 2018, San Francisco, United States, 18-10-27. https://doi.org/10.23919/PNC.2018.8579455

An effective CNN approach for vertebrae segmentation from 3D CT images. / Kuok, Chan Pang; Hsue, Jin Yuan; Shen, Ting Li; Huang, Bing Feng; Chen, Chi Yeh; Sun, Yung Nien.

Proceedings of the 2018 Pacific Neighborhood Consortium Annual Conference and Joint Meetings: Human Rights in Cyberspace, PNC 2018. ed. / Shih-Lung Shaw; Ta-Chien Chan; Ling-Jyh Chen. Institute of Electrical and Electronics Engineers Inc., 2018. p. 7-12 8579455.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - An effective CNN approach for vertebrae segmentation from 3D CT images

AU - Kuok, Chan Pang

AU - Hsue, Jin Yuan

AU - Shen, Ting Li

AU - Huang, Bing Feng

AU - Chen, Chi Yeh

AU - Sun, Yung Nien

PY - 2018/12/14

Y1 - 2018/12/14

N2 - Spine surgery is risky when treating spinal diseases or injuries. In order to decrease the damage of tissue and accelerate the recovery, minimally invasive surgery is strongly desired. Because of this, the preoperative planning and intraoperative guidance become very important. A 3D vertebrae model of the surgical site will be built for these purposes from the computer tomography (CT) images. The segmentation methods of vertebrae from the CT images for modeling are widely concerned. The conventional methods were satisfactory but they usually required prior knowledge of the vertebrae. Recently, the deep learning convolutional neural network (CNN) showed an outstanding performance on the end-to-end semantic image segmentation. Benefit from this, we propose an effective and simple convolution neural network based approach for vertebrae segmentation. Different from training and testing on large 3D data, the axial view images are used effectively in this study. A sequence of simple image processing steps are applied for the intervertebral plane detection. Five CT cases are used on this study and the segmentation results are highly accurate with average dice similarity coefficient over 0.95.

AB - Spine surgery is risky when treating spinal diseases or injuries. In order to decrease the damage of tissue and accelerate the recovery, minimally invasive surgery is strongly desired. Because of this, the preoperative planning and intraoperative guidance become very important. A 3D vertebrae model of the surgical site will be built for these purposes from the computer tomography (CT) images. The segmentation methods of vertebrae from the CT images for modeling are widely concerned. The conventional methods were satisfactory but they usually required prior knowledge of the vertebrae. Recently, the deep learning convolutional neural network (CNN) showed an outstanding performance on the end-to-end semantic image segmentation. Benefit from this, we propose an effective and simple convolution neural network based approach for vertebrae segmentation. Different from training and testing on large 3D data, the axial view images are used effectively in this study. A sequence of simple image processing steps are applied for the intervertebral plane detection. Five CT cases are used on this study and the segmentation results are highly accurate with average dice similarity coefficient over 0.95.

UR - http://www.scopus.com/inward/record.url?scp=85060889527&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85060889527&partnerID=8YFLogxK

U2 - 10.23919/PNC.2018.8579455

DO - 10.23919/PNC.2018.8579455

M3 - Conference contribution

AN - SCOPUS:85060889527

SP - 7

EP - 12

BT - Proceedings of the 2018 Pacific Neighborhood Consortium Annual Conference and Joint Meetings

A2 - Shaw, Shih-Lung

A2 - Chan, Ta-Chien

A2 - Chen, Ling-Jyh

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Kuok CP, Hsue JY, Shen TL, Huang BF, Chen CY, Sun YN. An effective CNN approach for vertebrae segmentation from 3D CT images. In Shaw S-L, Chan T-C, Chen L-J, editors, Proceedings of the 2018 Pacific Neighborhood Consortium Annual Conference and Joint Meetings: Human Rights in Cyberspace, PNC 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 7-12. 8579455 https://doi.org/10.23919/PNC.2018.8579455