Discriminating between metal and near-metal pixels in X-Ray CT images

Hanjae Park; Seho Bae; Seongbeak Yoon; Lei Li; Jixuan Liu; Juneho Yi

doi:10.1145/2701126.2701136

Discriminating between metal and near-metal pixels in X-Ray CT images

Hanjae Park
, Seho Bae
, Seongbeak Yoon
, Lei Li
, Jixuan Liu
, Juneho Yi

Department of Electronic and Electrical Engineering

Sungkyunkwan University

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

Abstract

This research presents a novel feature that is capable of significantly facilitating distinction between metal pixels and nonmetal pixels near metal in X-ray CT images. In order to create this feature, we exploit the Relevant Neighbor Area (RNA) idea originally developed for projection image inpainting. The feature distinctively distributes metal and near-metal pixels with opposite signs. Simple thresholding results on the histogram of the reconstructed image of the feature image shows the effectiveness of the feature proposed. The proposed feature, when combined with a segmentation algorithm, is expected to significantly contribute to the solution of metal area segmentation in X-ray CT images.

Original language	English
Title of host publication	ACM IMCOM 2015 - Proceedings
Publisher	Association for Computing Machinery, Inc
ISBN (Electronic)	9781450333771
DOIs	https://doi.org/10.1145/2701126.2701136
State	Published - 8 Jan 2015
Event	9th International Conference on Ubiquitous Information Management and Communication, ACM IMCOM 2015 - Bali, Indonesia Duration: 8 Jan 2015 → 10 Jan 2015

Publication series

Name	ACM IMCOM 2015 - Proceedings

Conference

Conference	9th International Conference on Ubiquitous Information Management and Communication, ACM IMCOM 2015
Country/Territory	Indonesia
City	Bali
Period	8/01/15 → 10/01/15

Keywords

Metal area segmentation
Metal artifacts reduction
Relevant neighbor area
X-ray CT image

Access to Document

10.1145/2701126.2701136

Cite this

@inproceedings{23efff63b590488997a33404b25752c3,

title = "Discriminating between metal and near-metal pixels in X-Ray CT images",

abstract = "This research presents a novel feature that is capable of significantly facilitating distinction between metal pixels and nonmetal pixels near metal in X-ray CT images. In order to create this feature, we exploit the Relevant Neighbor Area (RNA) idea originally developed for projection image inpainting. The feature distinctively distributes metal and near-metal pixels with opposite signs. Simple thresholding results on the histogram of the reconstructed image of the feature image shows the effectiveness of the feature proposed. The proposed feature, when combined with a segmentation algorithm, is expected to significantly contribute to the solution of metal area segmentation in X-ray CT images.",

keywords = "Metal area segmentation, Metal artifacts reduction, Relevant neighbor area, X-ray CT image",

author = "Hanjae Park and Seho Bae and Seongbeak Yoon and Lei Li and Jixuan Liu and Juneho Yi",

year = "2015",

month = jan,

day = "8",

doi = "10.1145/2701126.2701136",

language = "English",

series = "ACM IMCOM 2015 - Proceedings",

publisher = "Association for Computing Machinery, Inc",

booktitle = "ACM IMCOM 2015 - Proceedings",

note = "9th International Conference on Ubiquitous Information Management and Communication, ACM IMCOM 2015 ; Conference date: 08-01-2015 Through 10-01-2015",

}

Park, H, Bae, S, Yoon, S, Li, L, Liu, J & Yi, J 2015, Discriminating between metal and near-metal pixels in X-Ray CT images. in ACM IMCOM 2015 - Proceedings., a63, ACM IMCOM 2015 - Proceedings, Association for Computing Machinery, Inc, 9th International Conference on Ubiquitous Information Management and Communication, ACM IMCOM 2015, Bali, Indonesia, 8/01/15. https://doi.org/10.1145/2701126.2701136

TY - GEN

T1 - Discriminating between metal and near-metal pixels in X-Ray CT images

AU - Park, Hanjae

AU - Bae, Seho

AU - Yoon, Seongbeak

AU - Li, Lei

AU - Liu, Jixuan

AU - Yi, Juneho

PY - 2015/1/8

Y1 - 2015/1/8

N2 - This research presents a novel feature that is capable of significantly facilitating distinction between metal pixels and nonmetal pixels near metal in X-ray CT images. In order to create this feature, we exploit the Relevant Neighbor Area (RNA) idea originally developed for projection image inpainting. The feature distinctively distributes metal and near-metal pixels with opposite signs. Simple thresholding results on the histogram of the reconstructed image of the feature image shows the effectiveness of the feature proposed. The proposed feature, when combined with a segmentation algorithm, is expected to significantly contribute to the solution of metal area segmentation in X-ray CT images.

AB - This research presents a novel feature that is capable of significantly facilitating distinction between metal pixels and nonmetal pixels near metal in X-ray CT images. In order to create this feature, we exploit the Relevant Neighbor Area (RNA) idea originally developed for projection image inpainting. The feature distinctively distributes metal and near-metal pixels with opposite signs. Simple thresholding results on the histogram of the reconstructed image of the feature image shows the effectiveness of the feature proposed. The proposed feature, when combined with a segmentation algorithm, is expected to significantly contribute to the solution of metal area segmentation in X-ray CT images.

KW - Metal area segmentation

KW - Metal artifacts reduction

KW - Relevant neighbor area

KW - X-ray CT image

UR - https://www.scopus.com/pages/publications/84926212468

U2 - 10.1145/2701126.2701136

DO - 10.1145/2701126.2701136

M3 - Conference contribution

AN - SCOPUS:84926212468

T3 - ACM IMCOM 2015 - Proceedings

BT - ACM IMCOM 2015 - Proceedings

PB - Association for Computing Machinery, Inc

T2 - 9th International Conference on Ubiquitous Information Management and Communication, ACM IMCOM 2015

Y2 - 8 January 2015 through 10 January 2015

ER -

Discriminating between metal and near-metal pixels in X-Ray CT images

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this