Evaluation of segmentation algorithms for optical coherence tomography images of ovarian tissue

Travis W. Sawyer, Photini F.S. Rice, David M. Sawyer, Jennifer W. Koevary, Jennifer K Barton

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

1 Citation (Scopus)

Abstract

Ovarian cancer has the lowest survival rate among all gynecologic cancers due to predominantly late diagnosis. Early detection of ovarian cancer can increase 5-year survival rates from 40% up to 92%, yet no reliable early detection techniques exist. Optical coherence tomography (OCT) is an emerging technique that provides depthresolved, high-resolution images of biological tissue in real time and demonstrates great potential for imaging of ovarian tissue. Mouse models are crucial to quantitatively assess the diagnostic potential of OCT for ovarian cancer imaging; however, due to small organ size, the ovaries must rst be separated from the image background using the process of segmentation. Manual segmentation is time-intensive, as OCT yields three-dimensional data. Furthermore, speckle noise complicates OCT images, frustrating many processing techniques. While much work has investigated noise-reduction and automated segmentation for retinal OCT imaging, little has considered the application to the ovaries, which exhibit higher variance and inhomogeneity than the retina. To address these challenges, we evaluated a set of algorithms to segment OCT images of mouse ovaries. We examined ve preprocessing techniques and six segmentation algorithms. While all pre-processing methods improve segmentation, Gaussian filtering is most effective, showing an improvement of 32% ± 1.2%. Of the segmentation algorithms, active contours performs best, segmenting with an accuracy of 0.948 ± 0.012 compared with manual segmentation (1.0 being identical). Nonetheless, further optimization could lead to maximizing the performance for segmenting OCT images of the ovaries.

Original languageEnglish (US)
Title of host publicationDiagnosis and Treatment of Diseases in the Breast and Reproductive System IV
PublisherSPIE
Volume10472
ISBN (Electronic)9781510614291
DOIs
StatePublished - Jan 1 2018
EventDiagnosis and Treatment of Diseases in the Breast and Reproductive System IV 2018 - San Francisco, United States
Duration: Jan 27 2018Jan 28 2018

Other

OtherDiagnosis and Treatment of Diseases in the Breast and Reproductive System IV 2018
CountryUnited States
CitySan Francisco
Period1/27/181/28/18

Fingerprint

Optical tomography
Optical Coherence Tomography
tomography
ovaries
Tissue
evaluation
Ovary
cancer
Ovarian Neoplasms
preprocessing
Imaging techniques
mice
Organ Size
retina
Delayed Diagnosis
Image resolution
Speckle
Processing
Noise abatement
noise reduction

Keywords

  • image processing
  • image segmentation
  • optical coherence tomography
  • ovarian cancer

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Biomaterials
  • Atomic and Molecular Physics, and Optics
  • Radiology Nuclear Medicine and imaging

Cite this

Sawyer, T. W., Rice, P. F. S., Sawyer, D. M., Koevary, J. W., & Barton, J. K. (2018). Evaluation of segmentation algorithms for optical coherence tomography images of ovarian tissue. In Diagnosis and Treatment of Diseases in the Breast and Reproductive System IV (Vol. 10472). [1047204] SPIE. https://doi.org/10.1117/12.2283375

Evaluation of segmentation algorithms for optical coherence tomography images of ovarian tissue. / Sawyer, Travis W.; Rice, Photini F.S.; Sawyer, David M.; Koevary, Jennifer W.; Barton, Jennifer K.

Diagnosis and Treatment of Diseases in the Breast and Reproductive System IV. Vol. 10472 SPIE, 2018. 1047204.

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

Sawyer, TW, Rice, PFS, Sawyer, DM, Koevary, JW & Barton, JK 2018, Evaluation of segmentation algorithms for optical coherence tomography images of ovarian tissue. in Diagnosis and Treatment of Diseases in the Breast and Reproductive System IV. vol. 10472, 1047204, SPIE, Diagnosis and Treatment of Diseases in the Breast and Reproductive System IV 2018, San Francisco, United States, 1/27/18. https://doi.org/10.1117/12.2283375
Sawyer TW, Rice PFS, Sawyer DM, Koevary JW, Barton JK. Evaluation of segmentation algorithms for optical coherence tomography images of ovarian tissue. In Diagnosis and Treatment of Diseases in the Breast and Reproductive System IV. Vol. 10472. SPIE. 2018. 1047204 https://doi.org/10.1117/12.2283375
Sawyer, Travis W. ; Rice, Photini F.S. ; Sawyer, David M. ; Koevary, Jennifer W. ; Barton, Jennifer K. / Evaluation of segmentation algorithms for optical coherence tomography images of ovarian tissue. Diagnosis and Treatment of Diseases in the Breast and Reproductive System IV. Vol. 10472 SPIE, 2018.
@inproceedings{0eae8cdb70ee4e16bd15823f1df8a22a,
title = "Evaluation of segmentation algorithms for optical coherence tomography images of ovarian tissue",
abstract = "Ovarian cancer has the lowest survival rate among all gynecologic cancers due to predominantly late diagnosis. Early detection of ovarian cancer can increase 5-year survival rates from 40{\%} up to 92{\%}, yet no reliable early detection techniques exist. Optical coherence tomography (OCT) is an emerging technique that provides depthresolved, high-resolution images of biological tissue in real time and demonstrates great potential for imaging of ovarian tissue. Mouse models are crucial to quantitatively assess the diagnostic potential of OCT for ovarian cancer imaging; however, due to small organ size, the ovaries must rst be separated from the image background using the process of segmentation. Manual segmentation is time-intensive, as OCT yields three-dimensional data. Furthermore, speckle noise complicates OCT images, frustrating many processing techniques. While much work has investigated noise-reduction and automated segmentation for retinal OCT imaging, little has considered the application to the ovaries, which exhibit higher variance and inhomogeneity than the retina. To address these challenges, we evaluated a set of algorithms to segment OCT images of mouse ovaries. We examined ve preprocessing techniques and six segmentation algorithms. While all pre-processing methods improve segmentation, Gaussian filtering is most effective, showing an improvement of 32{\%} ± 1.2{\%}. Of the segmentation algorithms, active contours performs best, segmenting with an accuracy of 0.948 ± 0.012 compared with manual segmentation (1.0 being identical). Nonetheless, further optimization could lead to maximizing the performance for segmenting OCT images of the ovaries.",
keywords = "image processing, image segmentation, optical coherence tomography, ovarian cancer",
author = "Sawyer, {Travis W.} and Rice, {Photini F.S.} and Sawyer, {David M.} and Koevary, {Jennifer W.} and Barton, {Jennifer K}",
year = "2018",
month = "1",
day = "1",
doi = "10.1117/12.2283375",
language = "English (US)",
volume = "10472",
booktitle = "Diagnosis and Treatment of Diseases in the Breast and Reproductive System IV",
publisher = "SPIE",

}

TY - GEN

T1 - Evaluation of segmentation algorithms for optical coherence tomography images of ovarian tissue

AU - Sawyer, Travis W.

AU - Rice, Photini F.S.

AU - Sawyer, David M.

AU - Koevary, Jennifer W.

AU - Barton, Jennifer K

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Ovarian cancer has the lowest survival rate among all gynecologic cancers due to predominantly late diagnosis. Early detection of ovarian cancer can increase 5-year survival rates from 40% up to 92%, yet no reliable early detection techniques exist. Optical coherence tomography (OCT) is an emerging technique that provides depthresolved, high-resolution images of biological tissue in real time and demonstrates great potential for imaging of ovarian tissue. Mouse models are crucial to quantitatively assess the diagnostic potential of OCT for ovarian cancer imaging; however, due to small organ size, the ovaries must rst be separated from the image background using the process of segmentation. Manual segmentation is time-intensive, as OCT yields three-dimensional data. Furthermore, speckle noise complicates OCT images, frustrating many processing techniques. While much work has investigated noise-reduction and automated segmentation for retinal OCT imaging, little has considered the application to the ovaries, which exhibit higher variance and inhomogeneity than the retina. To address these challenges, we evaluated a set of algorithms to segment OCT images of mouse ovaries. We examined ve preprocessing techniques and six segmentation algorithms. While all pre-processing methods improve segmentation, Gaussian filtering is most effective, showing an improvement of 32% ± 1.2%. Of the segmentation algorithms, active contours performs best, segmenting with an accuracy of 0.948 ± 0.012 compared with manual segmentation (1.0 being identical). Nonetheless, further optimization could lead to maximizing the performance for segmenting OCT images of the ovaries.

AB - Ovarian cancer has the lowest survival rate among all gynecologic cancers due to predominantly late diagnosis. Early detection of ovarian cancer can increase 5-year survival rates from 40% up to 92%, yet no reliable early detection techniques exist. Optical coherence tomography (OCT) is an emerging technique that provides depthresolved, high-resolution images of biological tissue in real time and demonstrates great potential for imaging of ovarian tissue. Mouse models are crucial to quantitatively assess the diagnostic potential of OCT for ovarian cancer imaging; however, due to small organ size, the ovaries must rst be separated from the image background using the process of segmentation. Manual segmentation is time-intensive, as OCT yields three-dimensional data. Furthermore, speckle noise complicates OCT images, frustrating many processing techniques. While much work has investigated noise-reduction and automated segmentation for retinal OCT imaging, little has considered the application to the ovaries, which exhibit higher variance and inhomogeneity than the retina. To address these challenges, we evaluated a set of algorithms to segment OCT images of mouse ovaries. We examined ve preprocessing techniques and six segmentation algorithms. While all pre-processing methods improve segmentation, Gaussian filtering is most effective, showing an improvement of 32% ± 1.2%. Of the segmentation algorithms, active contours performs best, segmenting with an accuracy of 0.948 ± 0.012 compared with manual segmentation (1.0 being identical). Nonetheless, further optimization could lead to maximizing the performance for segmenting OCT images of the ovaries.

KW - image processing

KW - image segmentation

KW - optical coherence tomography

KW - ovarian cancer

UR - http://www.scopus.com/inward/record.url?scp=85047005061&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85047005061&partnerID=8YFLogxK

U2 - 10.1117/12.2283375

DO - 10.1117/12.2283375

M3 - Conference contribution

AN - SCOPUS:85047005061

VL - 10472

BT - Diagnosis and Treatment of Diseases in the Breast and Reproductive System IV

PB - SPIE

ER -