Computer-aided diagnosis for distinguishing between triple-negative breast cancer and fibroadenomas based on ultrasound texture features

Woo Kyung Moon, Yao Sian Huang, Chung Ming Lo, Chiun Sheng Huang, Min Sun Bae, Won Hwa Kim, Jeon Hor Chen, Ruey Feng Chang

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

Purpose: Triple-negative breast cancer (TNBC), an aggressive subtype, is frequently misclassified as fibroadenoma due to benign morphologic features on breast ultrasound (US). This study aims to develop a computer-aided diagnosis (CAD) system based on texture features for distinguishing between TNBC and benign fibroadenomas in US images. Methods: US images of 169 pathology-proven tumors (mean size, 1.65 cm; range, 0.7-3.0 cm) composed of 84 benign fibroadenomas and 85 TNBC tumors are used in this study. After a tumor is segmented out using the level-set method, morphological, conventional texture, and multiresolution gray-scale invariant texture feature sets are computed using a best-fitting ellipse, gray-level co-occurrence matrices, and the ranklet transform, respectively. The linear support vector machine with leave-one-out cross-validation schema is used as a classifier, and the diagnostic performance is assessed with receiver operating characteristic curve analysis. Results: The Az values of the morphology, conventional texture, and multiresolution gray-scale invariant texture feature sets are 0.8470 [95% confidence intervals (CIs), 0.7826-0.8973], 0.8542 (95% CI, 0.7911-0.9030), and 0.9695 (95% CI, 0.9376-0.9865), respectively. The Az of the CAD system based on the combined feature sets is 0.9702 (95% CI, 0.9334-0.9882). Conclusions: The CAD system based on texture features extracted via the ranklet transform may be useful for improving the ability to discriminate between TNBC and benign fibroadenomas.

Original languageEnglish
Pages (from-to)3024-3035
Number of pages12
JournalMedical Physics
Volume42
Issue number6
DOIs
Publication statusPublished - Jun 1 2015
Externally publishedYes

Fingerprint

Triple Negative Breast Neoplasms
Fibroadenoma
Confidence Intervals
ROC Curve
Neoplasms
Breast
Pathology
Breast Neoplasms

Keywords

  • breast cancer
  • fibroadenoma
  • gray-scale invariant features
  • ranklet transform
  • triple-negative breast cancer

ASJC Scopus subject areas

  • Biophysics
  • Radiology Nuclear Medicine and imaging

Cite this

Moon, W. K., Huang, Y. S., Lo, C. M., Huang, C. S., Bae, M. S., Kim, W. H., ... Chang, R. F. (2015). Computer-aided diagnosis for distinguishing between triple-negative breast cancer and fibroadenomas based on ultrasound texture features. Medical Physics, 42(6), 3024-3035. https://doi.org/10.1118/1.4921123

Computer-aided diagnosis for distinguishing between triple-negative breast cancer and fibroadenomas based on ultrasound texture features. / Moon, Woo Kyung; Huang, Yao Sian; Lo, Chung Ming; Huang, Chiun Sheng; Bae, Min Sun; Kim, Won Hwa; Chen, Jeon Hor; Chang, Ruey Feng.

In: Medical Physics, Vol. 42, No. 6, 01.06.2015, p. 3024-3035.

Research output: Contribution to journalArticle

Moon, Woo Kyung ; Huang, Yao Sian ; Lo, Chung Ming ; Huang, Chiun Sheng ; Bae, Min Sun ; Kim, Won Hwa ; Chen, Jeon Hor ; Chang, Ruey Feng. / Computer-aided diagnosis for distinguishing between triple-negative breast cancer and fibroadenomas based on ultrasound texture features. In: Medical Physics. 2015 ; Vol. 42, No. 6. pp. 3024-3035.
@article{7a6b17c7b92f4489be0fadc8bca932db,
title = "Computer-aided diagnosis for distinguishing between triple-negative breast cancer and fibroadenomas based on ultrasound texture features",
abstract = "Purpose: Triple-negative breast cancer (TNBC), an aggressive subtype, is frequently misclassified as fibroadenoma due to benign morphologic features on breast ultrasound (US). This study aims to develop a computer-aided diagnosis (CAD) system based on texture features for distinguishing between TNBC and benign fibroadenomas in US images. Methods: US images of 169 pathology-proven tumors (mean size, 1.65 cm; range, 0.7-3.0 cm) composed of 84 benign fibroadenomas and 85 TNBC tumors are used in this study. After a tumor is segmented out using the level-set method, morphological, conventional texture, and multiresolution gray-scale invariant texture feature sets are computed using a best-fitting ellipse, gray-level co-occurrence matrices, and the ranklet transform, respectively. The linear support vector machine with leave-one-out cross-validation schema is used as a classifier, and the diagnostic performance is assessed with receiver operating characteristic curve analysis. Results: The Az values of the morphology, conventional texture, and multiresolution gray-scale invariant texture feature sets are 0.8470 [95{\%} confidence intervals (CIs), 0.7826-0.8973], 0.8542 (95{\%} CI, 0.7911-0.9030), and 0.9695 (95{\%} CI, 0.9376-0.9865), respectively. The Az of the CAD system based on the combined feature sets is 0.9702 (95{\%} CI, 0.9334-0.9882). Conclusions: The CAD system based on texture features extracted via the ranklet transform may be useful for improving the ability to discriminate between TNBC and benign fibroadenomas.",
keywords = "breast cancer, fibroadenoma, gray-scale invariant features, ranklet transform, triple-negative breast cancer",
author = "Moon, {Woo Kyung} and Huang, {Yao Sian} and Lo, {Chung Ming} and Huang, {Chiun Sheng} and Bae, {Min Sun} and Kim, {Won Hwa} and Chen, {Jeon Hor} and Chang, {Ruey Feng}",
year = "2015",
month = "6",
day = "1",
doi = "10.1118/1.4921123",
language = "English",
volume = "42",
pages = "3024--3035",
journal = "Medical Physics",
issn = "0094-2405",
publisher = "AAPM - American Association of Physicists in Medicine",
number = "6",

}

TY - JOUR

T1 - Computer-aided diagnosis for distinguishing between triple-negative breast cancer and fibroadenomas based on ultrasound texture features

AU - Moon, Woo Kyung

AU - Huang, Yao Sian

AU - Lo, Chung Ming

AU - Huang, Chiun Sheng

AU - Bae, Min Sun

AU - Kim, Won Hwa

AU - Chen, Jeon Hor

AU - Chang, Ruey Feng

PY - 2015/6/1

Y1 - 2015/6/1

N2 - Purpose: Triple-negative breast cancer (TNBC), an aggressive subtype, is frequently misclassified as fibroadenoma due to benign morphologic features on breast ultrasound (US). This study aims to develop a computer-aided diagnosis (CAD) system based on texture features for distinguishing between TNBC and benign fibroadenomas in US images. Methods: US images of 169 pathology-proven tumors (mean size, 1.65 cm; range, 0.7-3.0 cm) composed of 84 benign fibroadenomas and 85 TNBC tumors are used in this study. After a tumor is segmented out using the level-set method, morphological, conventional texture, and multiresolution gray-scale invariant texture feature sets are computed using a best-fitting ellipse, gray-level co-occurrence matrices, and the ranklet transform, respectively. The linear support vector machine with leave-one-out cross-validation schema is used as a classifier, and the diagnostic performance is assessed with receiver operating characteristic curve analysis. Results: The Az values of the morphology, conventional texture, and multiresolution gray-scale invariant texture feature sets are 0.8470 [95% confidence intervals (CIs), 0.7826-0.8973], 0.8542 (95% CI, 0.7911-0.9030), and 0.9695 (95% CI, 0.9376-0.9865), respectively. The Az of the CAD system based on the combined feature sets is 0.9702 (95% CI, 0.9334-0.9882). Conclusions: The CAD system based on texture features extracted via the ranklet transform may be useful for improving the ability to discriminate between TNBC and benign fibroadenomas.

AB - Purpose: Triple-negative breast cancer (TNBC), an aggressive subtype, is frequently misclassified as fibroadenoma due to benign morphologic features on breast ultrasound (US). This study aims to develop a computer-aided diagnosis (CAD) system based on texture features for distinguishing between TNBC and benign fibroadenomas in US images. Methods: US images of 169 pathology-proven tumors (mean size, 1.65 cm; range, 0.7-3.0 cm) composed of 84 benign fibroadenomas and 85 TNBC tumors are used in this study. After a tumor is segmented out using the level-set method, morphological, conventional texture, and multiresolution gray-scale invariant texture feature sets are computed using a best-fitting ellipse, gray-level co-occurrence matrices, and the ranklet transform, respectively. The linear support vector machine with leave-one-out cross-validation schema is used as a classifier, and the diagnostic performance is assessed with receiver operating characteristic curve analysis. Results: The Az values of the morphology, conventional texture, and multiresolution gray-scale invariant texture feature sets are 0.8470 [95% confidence intervals (CIs), 0.7826-0.8973], 0.8542 (95% CI, 0.7911-0.9030), and 0.9695 (95% CI, 0.9376-0.9865), respectively. The Az of the CAD system based on the combined feature sets is 0.9702 (95% CI, 0.9334-0.9882). Conclusions: The CAD system based on texture features extracted via the ranklet transform may be useful for improving the ability to discriminate between TNBC and benign fibroadenomas.

KW - breast cancer

KW - fibroadenoma

KW - gray-scale invariant features

KW - ranklet transform

KW - triple-negative breast cancer

UR - http://www.scopus.com/inward/record.url?scp=84931266806&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84931266806&partnerID=8YFLogxK

U2 - 10.1118/1.4921123

DO - 10.1118/1.4921123

M3 - Article

VL - 42

SP - 3024

EP - 3035

JO - Medical Physics

JF - Medical Physics

SN - 0094-2405

IS - 6

ER -