Discriminative sparse representations for cervigram image segmentation

Shaoting Zhang, Junzhou Huang, Dimitris Metaxas, Wei Wang, Sharon Xiaolei Huang

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

22 Citations (Scopus)

Abstract

This paper presents an algorithm using discriminative sparse representations to segment tissues in optical images of the uterine cervix. Because of the large variations in the image appearance caused by the changing of illumination and specular reflection, the different classes of color and texture features in optical images are often overlapped with each other. Using sparse representations they can be transformed to higher dimension with sparse constraints and become more linearly separated. Different from the previous reconstructive sparse representation, the discriminative method considers positive and negative samples simultaneously, which means that these generated dictionaries can be discriminative and perform better for their own classes but worse for the others. New data can be reconstructed from its sparse representations and positive and/or negative dictionaries. Classification can be achieved based on comparing the reconstructive errors. In the experiments we used our method to automatically segment the biomarker AcetoWhite (AW) regions in an archive of the uterine cervix. Compared with the other general methods including SVM, nearest neighbor and reconstructive sparse representations, our approach showed higher sensitivity and specificity.

Original languageEnglish (US)
Title of host publication2010 7th IEEE International Symposium on Biomedical Imaging
Subtitle of host publicationFrom Nano to Macro, ISBI 2010 - Proceedings
PublisherIEEE Computer Society
Pages133-136
Number of pages4
ISBN (Print)9781424441266
DOIs
StatePublished - Jan 1 2010
Event7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Rotterdam, Netherlands
Duration: Apr 14 2010Apr 17 2010

Publication series

Name2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings

Conference

Conference7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010
CountryNetherlands
CityRotterdam
Period4/14/104/17/10

Fingerprint

Glossaries
Image segmentation
Cervix Uteri
Biomarkers
Textures
Lighting
Tissue
Color
Sensitivity and Specificity
Experiments

All Science Journal Classification (ASJC) codes

  • Biomedical Engineering
  • Radiology Nuclear Medicine and imaging

Cite this

Zhang, S., Huang, J., Metaxas, D., Wang, W., & Huang, S. X. (2010). Discriminative sparse representations for cervigram image segmentation. In 2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings (pp. 133-136). [5490397] (2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings). IEEE Computer Society. https://doi.org/10.1109/ISBI.2010.5490397
Zhang, Shaoting ; Huang, Junzhou ; Metaxas, Dimitris ; Wang, Wei ; Huang, Sharon Xiaolei. / Discriminative sparse representations for cervigram image segmentation. 2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings. IEEE Computer Society, 2010. pp. 133-136 (2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings).
@inproceedings{6e68ce2c099f4c278e1b858664c4c855,
title = "Discriminative sparse representations for cervigram image segmentation",
abstract = "This paper presents an algorithm using discriminative sparse representations to segment tissues in optical images of the uterine cervix. Because of the large variations in the image appearance caused by the changing of illumination and specular reflection, the different classes of color and texture features in optical images are often overlapped with each other. Using sparse representations they can be transformed to higher dimension with sparse constraints and become more linearly separated. Different from the previous reconstructive sparse representation, the discriminative method considers positive and negative samples simultaneously, which means that these generated dictionaries can be discriminative and perform better for their own classes but worse for the others. New data can be reconstructed from its sparse representations and positive and/or negative dictionaries. Classification can be achieved based on comparing the reconstructive errors. In the experiments we used our method to automatically segment the biomarker AcetoWhite (AW) regions in an archive of the uterine cervix. Compared with the other general methods including SVM, nearest neighbor and reconstructive sparse representations, our approach showed higher sensitivity and specificity.",
author = "Shaoting Zhang and Junzhou Huang and Dimitris Metaxas and Wei Wang and Huang, {Sharon Xiaolei}",
year = "2010",
month = "1",
day = "1",
doi = "10.1109/ISBI.2010.5490397",
language = "English (US)",
isbn = "9781424441266",
series = "2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings",
publisher = "IEEE Computer Society",
pages = "133--136",
booktitle = "2010 7th IEEE International Symposium on Biomedical Imaging",
address = "United States",

}

Zhang, S, Huang, J, Metaxas, D, Wang, W & Huang, SX 2010, Discriminative sparse representations for cervigram image segmentation. in 2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings., 5490397, 2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings, IEEE Computer Society, pp. 133-136, 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010, Rotterdam, Netherlands, 4/14/10. https://doi.org/10.1109/ISBI.2010.5490397

Discriminative sparse representations for cervigram image segmentation. / Zhang, Shaoting; Huang, Junzhou; Metaxas, Dimitris; Wang, Wei; Huang, Sharon Xiaolei.

2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings. IEEE Computer Society, 2010. p. 133-136 5490397 (2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings).

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

TY - GEN

T1 - Discriminative sparse representations for cervigram image segmentation

AU - Zhang, Shaoting

AU - Huang, Junzhou

AU - Metaxas, Dimitris

AU - Wang, Wei

AU - Huang, Sharon Xiaolei

PY - 2010/1/1

Y1 - 2010/1/1

N2 - This paper presents an algorithm using discriminative sparse representations to segment tissues in optical images of the uterine cervix. Because of the large variations in the image appearance caused by the changing of illumination and specular reflection, the different classes of color and texture features in optical images are often overlapped with each other. Using sparse representations they can be transformed to higher dimension with sparse constraints and become more linearly separated. Different from the previous reconstructive sparse representation, the discriminative method considers positive and negative samples simultaneously, which means that these generated dictionaries can be discriminative and perform better for their own classes but worse for the others. New data can be reconstructed from its sparse representations and positive and/or negative dictionaries. Classification can be achieved based on comparing the reconstructive errors. In the experiments we used our method to automatically segment the biomarker AcetoWhite (AW) regions in an archive of the uterine cervix. Compared with the other general methods including SVM, nearest neighbor and reconstructive sparse representations, our approach showed higher sensitivity and specificity.

AB - This paper presents an algorithm using discriminative sparse representations to segment tissues in optical images of the uterine cervix. Because of the large variations in the image appearance caused by the changing of illumination and specular reflection, the different classes of color and texture features in optical images are often overlapped with each other. Using sparse representations they can be transformed to higher dimension with sparse constraints and become more linearly separated. Different from the previous reconstructive sparse representation, the discriminative method considers positive and negative samples simultaneously, which means that these generated dictionaries can be discriminative and perform better for their own classes but worse for the others. New data can be reconstructed from its sparse representations and positive and/or negative dictionaries. Classification can be achieved based on comparing the reconstructive errors. In the experiments we used our method to automatically segment the biomarker AcetoWhite (AW) regions in an archive of the uterine cervix. Compared with the other general methods including SVM, nearest neighbor and reconstructive sparse representations, our approach showed higher sensitivity and specificity.

UR - http://www.scopus.com/inward/record.url?scp=77955202718&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77955202718&partnerID=8YFLogxK

U2 - 10.1109/ISBI.2010.5490397

DO - 10.1109/ISBI.2010.5490397

M3 - Conference contribution

SN - 9781424441266

T3 - 2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings

SP - 133

EP - 136

BT - 2010 7th IEEE International Symposium on Biomedical Imaging

PB - IEEE Computer Society

ER -

Zhang S, Huang J, Metaxas D, Wang W, Huang SX. Discriminative sparse representations for cervigram image segmentation. In 2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings. IEEE Computer Society. 2010. p. 133-136. 5490397. (2010 7th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2010 - Proceedings). https://doi.org/10.1109/ISBI.2010.5490397