Automated texture-based identification of ovarian cancer in confocal microendoscope images

Saurabh Srivastava, Jeffrey J Rodriguez, Andrew R Rouse, Molly A. Brewer, Arthur F Gmitro

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

7 Citations (Scopus)

Abstract

The fluorescence confocal microendoscope provides high-resolution, in-vivo imaging of cellular pathology during optical biopsy. There are indications that the examination of human ovaries with this instrument has diagnostic implications for the early detection of ovarian cancer. The purpose of this study was to develop a computer-aided system to facilitate the identification of ovarian cancer from digital images captured with the confocal microendoscope system. To achieve this goal, we modeled the cellular-level structure present in these images as texture and extracted features based on first-order statistics, spatial gray-level dependence matrices, and spatial-frequency content. Selection of the best features for classification was performed using traditional feature selection techniques including stepwise discriminant analysis, forward sequential search, a non-parametric method, principal component analysis, and a heuristic technique that combines the results of these methods. The best set of features selected was used for classification, and performance of various machine classifiers was compared by analyzing the areas under their receiver operating characteristic curves. The results show that it is possible to automatically identify patients with ovarian cancer based on texture features extracted from confocal microendoscope images and that the machine performance is superior to that of the human observer.

Original languageEnglish (US)
Title of host publicationProgress in Biomedical Optics and Imaging - Proceedings of SPIE
EditorsJ.-A. Conchello, C.J. Cogswell, T. Wilson
Pages42-52
Number of pages11
Volume5701
DOIs
StatePublished - 2005
EventThree- Dimensional and Multidomensional Microscopy: Image Acquisition and Processing XII - San Jose, CA, United States
Duration: Jan 25 2005Jan 27 2005

Other

OtherThree- Dimensional and Multidomensional Microscopy: Image Acquisition and Processing XII
CountryUnited States
CitySan Jose, CA
Period1/25/051/27/05

Fingerprint

Textures
Biopsy
Discriminant analysis
Pathology
Principal component analysis
Feature extraction
Classifiers
Fluorescence
Statistics
Imaging techniques

Keywords

  • Automated classification
  • Confocal microendoscope
  • Ovarian cancer
  • Pattern recognition
  • Texture analysis

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Srivastava, S., Rodriguez, J. J., Rouse, A. R., Brewer, M. A., & Gmitro, A. F. (2005). Automated texture-based identification of ovarian cancer in confocal microendoscope images. In J-A. Conchello, C. J. Cogswell, & T. Wilson (Eds.), Progress in Biomedical Optics and Imaging - Proceedings of SPIE (Vol. 5701, pp. 42-52). [07] https://doi.org/10.1117/12.590592

Automated texture-based identification of ovarian cancer in confocal microendoscope images. / Srivastava, Saurabh; Rodriguez, Jeffrey J; Rouse, Andrew R; Brewer, Molly A.; Gmitro, Arthur F.

Progress in Biomedical Optics and Imaging - Proceedings of SPIE. ed. / J.-A. Conchello; C.J. Cogswell; T. Wilson. Vol. 5701 2005. p. 42-52 07.

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

Srivastava, S, Rodriguez, JJ, Rouse, AR, Brewer, MA & Gmitro, AF 2005, Automated texture-based identification of ovarian cancer in confocal microendoscope images. in J-A Conchello, CJ Cogswell & T Wilson (eds), Progress in Biomedical Optics and Imaging - Proceedings of SPIE. vol. 5701, 07, pp. 42-52, Three- Dimensional and Multidomensional Microscopy: Image Acquisition and Processing XII, San Jose, CA, United States, 1/25/05. https://doi.org/10.1117/12.590592
Srivastava S, Rodriguez JJ, Rouse AR, Brewer MA, Gmitro AF. Automated texture-based identification of ovarian cancer in confocal microendoscope images. In Conchello J-A, Cogswell CJ, Wilson T, editors, Progress in Biomedical Optics and Imaging - Proceedings of SPIE. Vol. 5701. 2005. p. 42-52. 07 https://doi.org/10.1117/12.590592
Srivastava, Saurabh ; Rodriguez, Jeffrey J ; Rouse, Andrew R ; Brewer, Molly A. ; Gmitro, Arthur F. / Automated texture-based identification of ovarian cancer in confocal microendoscope images. Progress in Biomedical Optics and Imaging - Proceedings of SPIE. editor / J.-A. Conchello ; C.J. Cogswell ; T. Wilson. Vol. 5701 2005. pp. 42-52
@inproceedings{59180b276ae547b4b08817cd02b83459,
title = "Automated texture-based identification of ovarian cancer in confocal microendoscope images",
abstract = "The fluorescence confocal microendoscope provides high-resolution, in-vivo imaging of cellular pathology during optical biopsy. There are indications that the examination of human ovaries with this instrument has diagnostic implications for the early detection of ovarian cancer. The purpose of this study was to develop a computer-aided system to facilitate the identification of ovarian cancer from digital images captured with the confocal microendoscope system. To achieve this goal, we modeled the cellular-level structure present in these images as texture and extracted features based on first-order statistics, spatial gray-level dependence matrices, and spatial-frequency content. Selection of the best features for classification was performed using traditional feature selection techniques including stepwise discriminant analysis, forward sequential search, a non-parametric method, principal component analysis, and a heuristic technique that combines the results of these methods. The best set of features selected was used for classification, and performance of various machine classifiers was compared by analyzing the areas under their receiver operating characteristic curves. The results show that it is possible to automatically identify patients with ovarian cancer based on texture features extracted from confocal microendoscope images and that the machine performance is superior to that of the human observer.",
keywords = "Automated classification, Confocal microendoscope, Ovarian cancer, Pattern recognition, Texture analysis",
author = "Saurabh Srivastava and Rodriguez, {Jeffrey J} and Rouse, {Andrew R} and Brewer, {Molly A.} and Gmitro, {Arthur F}",
year = "2005",
doi = "10.1117/12.590592",
language = "English (US)",
volume = "5701",
pages = "42--52",
editor = "J.-A. Conchello and C.J. Cogswell and T. Wilson",
booktitle = "Progress in Biomedical Optics and Imaging - Proceedings of SPIE",

}

TY - GEN

T1 - Automated texture-based identification of ovarian cancer in confocal microendoscope images

AU - Srivastava, Saurabh

AU - Rodriguez, Jeffrey J

AU - Rouse, Andrew R

AU - Brewer, Molly A.

AU - Gmitro, Arthur F

PY - 2005

Y1 - 2005

N2 - The fluorescence confocal microendoscope provides high-resolution, in-vivo imaging of cellular pathology during optical biopsy. There are indications that the examination of human ovaries with this instrument has diagnostic implications for the early detection of ovarian cancer. The purpose of this study was to develop a computer-aided system to facilitate the identification of ovarian cancer from digital images captured with the confocal microendoscope system. To achieve this goal, we modeled the cellular-level structure present in these images as texture and extracted features based on first-order statistics, spatial gray-level dependence matrices, and spatial-frequency content. Selection of the best features for classification was performed using traditional feature selection techniques including stepwise discriminant analysis, forward sequential search, a non-parametric method, principal component analysis, and a heuristic technique that combines the results of these methods. The best set of features selected was used for classification, and performance of various machine classifiers was compared by analyzing the areas under their receiver operating characteristic curves. The results show that it is possible to automatically identify patients with ovarian cancer based on texture features extracted from confocal microendoscope images and that the machine performance is superior to that of the human observer.

AB - The fluorescence confocal microendoscope provides high-resolution, in-vivo imaging of cellular pathology during optical biopsy. There are indications that the examination of human ovaries with this instrument has diagnostic implications for the early detection of ovarian cancer. The purpose of this study was to develop a computer-aided system to facilitate the identification of ovarian cancer from digital images captured with the confocal microendoscope system. To achieve this goal, we modeled the cellular-level structure present in these images as texture and extracted features based on first-order statistics, spatial gray-level dependence matrices, and spatial-frequency content. Selection of the best features for classification was performed using traditional feature selection techniques including stepwise discriminant analysis, forward sequential search, a non-parametric method, principal component analysis, and a heuristic technique that combines the results of these methods. The best set of features selected was used for classification, and performance of various machine classifiers was compared by analyzing the areas under their receiver operating characteristic curves. The results show that it is possible to automatically identify patients with ovarian cancer based on texture features extracted from confocal microendoscope images and that the machine performance is superior to that of the human observer.

KW - Automated classification

KW - Confocal microendoscope

KW - Ovarian cancer

KW - Pattern recognition

KW - Texture analysis

UR - http://www.scopus.com/inward/record.url?scp=21844441680&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=21844441680&partnerID=8YFLogxK

U2 - 10.1117/12.590592

DO - 10.1117/12.590592

M3 - Conference contribution

VL - 5701

SP - 42

EP - 52

BT - Progress in Biomedical Optics and Imaging - Proceedings of SPIE

A2 - Conchello, J.-A.

A2 - Cogswell, C.J.

A2 - Wilson, T.

ER -