A novel approach based on joint optimization of alignment and statistical surface representation with wavelet transform for CBCT segmentation

Yu Bing Chang; Peng Yuan; Tai Hong Kuo; Zixiang Xiong; Jaime Gateno; James J. Xia; Xiaobo Zhou

doi:10.1007/978-3-642-38342-7_5

A novel approach based on joint optimization of alignment and statistical surface representation with wavelet transform for CBCT segmentation

Yu Bing Chang, Peng Yuan, Tai Hong Kuo, Zixiang Xiong, Jaime Gateno, James J. Xia, Xiaobo Zhou

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

Abstract

Cone-beam computed tomography (CBCT) can provide true 3D information of anatomical structures, with advantages of much thinner slice thickness and significantly lowered effective dose of radiation. However, CBCT images are extremely low contract and noisy. It is very difficult to segment thin bones. It usually takes 4-5 hours to manually segment a set of CBCT data. To this end, we developed a novel approach based on the joint optimization of alignment and statistical surface representation with wavelet transform for segmentation of CBCT images. It included two main steps: customized wavelet base initialization (CWBI) and base invariant wavelet active shape model (BIWASM). We validated our approach with others by comparing the surface deviation between segmented shape to the ground truth. The results showed that our approach outperformed the others in accuracy and computing time.

Original language	English
Title of host publication	Computational Intelligence Methods for Bioinformatics and Biostatistics
Subtitle of host publication	9th International Meeting, CIBB 2012, Revised Selected Papers
Publisher	Springer
Pages	48-56
Number of pages	9
Volume	7845 LNBI
ISBN (Print)	9783642383410
DOIs	https://doi.org/10.1007/978-3-642-38342-7_5
Publication status	Published - 2013
Externally published	Yes
Event	International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics (CIBB) 2012 - Houston, United States of America Duration: 12 Jul 2012 → 14 Jul 2012 Conference number: 9th

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	7845 LNBI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics (CIBB) 2012
Abbreviated title	CIBB 2012
Country/Territory	United States of America
City	Houston
Period	12/07/12 → 14/07/12

Keywords

Cone-beam CT
Segmentation
Statistical Shape Model

Access to Document

10.1007/978-3-642-38342-7_5

Cite this

Chang, Y. B., Yuan, P., Kuo, T. H., Xiong, Z., Gateno, J., Xia, J. J., & Zhou, X. (2013). A novel approach based on joint optimization of alignment and statistical surface representation with wavelet transform for CBCT segmentation. In Computational Intelligence Methods for Bioinformatics and Biostatistics: 9th International Meeting, CIBB 2012, Revised Selected Papers (Vol. 7845 LNBI, pp. 48-56). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7845 LNBI). Springer. https://doi.org/10.1007/978-3-642-38342-7_5

Chang, Yu Bing ; Yuan, Peng ; Kuo, Tai Hong et al. / A novel approach based on joint optimization of alignment and statistical surface representation with wavelet transform for CBCT segmentation. Computational Intelligence Methods for Bioinformatics and Biostatistics: 9th International Meeting, CIBB 2012, Revised Selected Papers. Vol. 7845 LNBI Springer, 2013. pp. 48-56 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{0cc59a25c55f46ceb41c04c0d22d8d9d,

title = "A novel approach based on joint optimization of alignment and statistical surface representation with wavelet transform for CBCT segmentation",

abstract = "Cone-beam computed tomography (CBCT) can provide true 3D information of anatomical structures, with advantages of much thinner slice thickness and significantly lowered effective dose of radiation. However, CBCT images are extremely low contract and noisy. It is very difficult to segment thin bones. It usually takes 4-5 hours to manually segment a set of CBCT data. To this end, we developed a novel approach based on the joint optimization of alignment and statistical surface representation with wavelet transform for segmentation of CBCT images. It included two main steps: customized wavelet base initialization (CWBI) and base invariant wavelet active shape model (BIWASM). We validated our approach with others by comparing the surface deviation between segmented shape to the ground truth. The results showed that our approach outperformed the others in accuracy and computing time.",

keywords = "Cone-beam CT, Segmentation, Statistical Shape Model",

author = "Chang, {Yu Bing} and Peng Yuan and Kuo, {Tai Hong} and Zixiang Xiong and Jaime Gateno and Xia, {James J.} and Xiaobo Zhou",

year = "2013",

doi = "10.1007/978-3-642-38342-7_5",

language = "English",

isbn = "9783642383410",

volume = "7845 LNBI",

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

publisher = "Springer",

pages = "48--56",

booktitle = "Computational Intelligence Methods for Bioinformatics and Biostatistics",

note = "International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics (CIBB) 2012, CIBB 2012 ; Conference date: 12-07-2012 Through 14-07-2012",

}

Chang, YB, Yuan, P, Kuo, TH, Xiong, Z, Gateno, J, Xia, JJ & Zhou, X 2013, A novel approach based on joint optimization of alignment and statistical surface representation with wavelet transform for CBCT segmentation. in Computational Intelligence Methods for Bioinformatics and Biostatistics: 9th International Meeting, CIBB 2012, Revised Selected Papers. vol. 7845 LNBI, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7845 LNBI, Springer, pp. 48-56, International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics (CIBB) 2012, Houston, Texas, United States of America, 12/07/12. https://doi.org/10.1007/978-3-642-38342-7_5

A novel approach based on joint optimization of alignment and statistical surface representation with wavelet transform for CBCT segmentation. / Chang, Yu Bing; Yuan, Peng; Kuo, Tai Hong et al.
Computational Intelligence Methods for Bioinformatics and Biostatistics: 9th International Meeting, CIBB 2012, Revised Selected Papers. Vol. 7845 LNBI Springer, 2013. p. 48-56 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7845 LNBI).

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

TY - GEN

T1 - A novel approach based on joint optimization of alignment and statistical surface representation with wavelet transform for CBCT segmentation

AU - Chang, Yu Bing

AU - Yuan, Peng

AU - Kuo, Tai Hong

AU - Xiong, Zixiang

AU - Gateno, Jaime

AU - Xia, James J.

AU - Zhou, Xiaobo

N1 - Conference code: 9th

PY - 2013

Y1 - 2013

N2 - Cone-beam computed tomography (CBCT) can provide true 3D information of anatomical structures, with advantages of much thinner slice thickness and significantly lowered effective dose of radiation. However, CBCT images are extremely low contract and noisy. It is very difficult to segment thin bones. It usually takes 4-5 hours to manually segment a set of CBCT data. To this end, we developed a novel approach based on the joint optimization of alignment and statistical surface representation with wavelet transform for segmentation of CBCT images. It included two main steps: customized wavelet base initialization (CWBI) and base invariant wavelet active shape model (BIWASM). We validated our approach with others by comparing the surface deviation between segmented shape to the ground truth. The results showed that our approach outperformed the others in accuracy and computing time.

AB - Cone-beam computed tomography (CBCT) can provide true 3D information of anatomical structures, with advantages of much thinner slice thickness and significantly lowered effective dose of radiation. However, CBCT images are extremely low contract and noisy. It is very difficult to segment thin bones. It usually takes 4-5 hours to manually segment a set of CBCT data. To this end, we developed a novel approach based on the joint optimization of alignment and statistical surface representation with wavelet transform for segmentation of CBCT images. It included two main steps: customized wavelet base initialization (CWBI) and base invariant wavelet active shape model (BIWASM). We validated our approach with others by comparing the surface deviation between segmented shape to the ground truth. The results showed that our approach outperformed the others in accuracy and computing time.

KW - Cone-beam CT

KW - Segmentation

KW - Statistical Shape Model

UR - http://www.scopus.com/inward/record.url?scp=84883395612&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-38342-7_5

DO - 10.1007/978-3-642-38342-7_5

M3 - Conference Paper

AN - SCOPUS:84883395612

SN - 9783642383410

VL - 7845 LNBI

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

SP - 48

EP - 56

BT - Computational Intelligence Methods for Bioinformatics and Biostatistics

PB - Springer

T2 - International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics (CIBB) 2012

Y2 - 12 July 2012 through 14 July 2012

ER -

Chang YB, Yuan P, Kuo TH, Xiong Z, Gateno J, Xia JJ et al. A novel approach based on joint optimization of alignment and statistical surface representation with wavelet transform for CBCT segmentation. In Computational Intelligence Methods for Bioinformatics and Biostatistics: 9th International Meeting, CIBB 2012, Revised Selected Papers. Vol. 7845 LNBI. Springer. 2013. p. 48-56. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-38342-7_5

A novel approach based on joint optimization of alignment and statistical surface representation with wavelet transform for CBCT segmentation

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Cite this