Evaluation of Computer-aided Diagnosis on a Large Clinical Full-field Digital Mammographic Dataset

Hui Li, Maryellen L. Giger, Yading Yuan, Weijie Chen, Karla Horsch, Li Lan, Andrew R. Jamieson, Charlene A. Sennett, Sanaz A. Jansen

Research output: Contribution to journalArticle

29 Citations (Scopus)

Abstract

Rationale and Objectives: To convert and optimize our previously developed computerized analysis methods for use with images from full-field digital mammography (FFDM) for breast mass classification to aid in the diagnosis of breast cancer. Materials and Methods: An institutional review board approved protocol was obtained, with waiver of consent for retrospective use of mammograms and pathology data. Seven hundred thirty-nine FFDM images, which contained 287 biopsy-proven breast mass lesions, of which 148 lesions were malignant and 139 lesions were benign, were retrospectively collected. Lesion margins were delineated by an expert breast radiologist and were used as the truth for lesion-segmentation evaluation. Our computerized image analysis method consisted of several steps: 1) identified lesions were automatically extracted from the parenchymal background using computerized segmentation methods; 2) a set of image characteristics (mathematic descriptors) were automatically extracted from image data of the lesions and surrounding tissues; and 3) selected features were merged into an estimate of the probability of malignancy using a Bayesian artificial neural network classifier. Performance of the analyses was evaluated at various stages of the conversion using receiver-operating characteristic analysis. Results: An area under the curve value of 0.81 was obtained in the task of distinguishing between malignant and benign mass lesions in a round-robin by case evaluation on the entire FFDM dataset. We failed to show a statistically significant difference (P = .83) compared to results from our previous study in which the computerized classification was performed on digitized screen-film mammograms. Conclusions: Our computerized analysis methods developed on digitized screen-film mammography can be converted for use with FFDM. Results show that the computerized analysis methods for the diagnosis of breast mass lesions on FFDM are promising, and can potentially be used to aid clinicians in the diagnostic interpretation of FFDM.

Original languageEnglish (US)
Pages (from-to)1437-1445
Number of pages9
JournalAcademic radiology
Volume15
Issue number11
DOIs
StatePublished - Nov 1 2008
Externally publishedYes

Fingerprint

Mammography
Breast
Songbirds
Mathematics
Research Ethics Committees
Datasets
ROC Curve
Area Under Curve
Pathology
Breast Neoplasms
Biopsy
Neoplasms

Keywords

  • breast mass classification
  • Computer-aided diagnosis
  • full-field digital mammography

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Cite this

Li, H., Giger, M. L., Yuan, Y., Chen, W., Horsch, K., Lan, L., ... Jansen, S. A. (2008). Evaluation of Computer-aided Diagnosis on a Large Clinical Full-field Digital Mammographic Dataset. Academic radiology, 15(11), 1437-1445. https://doi.org/10.1016/j.acra.2008.05.004

Evaluation of Computer-aided Diagnosis on a Large Clinical Full-field Digital Mammographic Dataset. / Li, Hui; Giger, Maryellen L.; Yuan, Yading; Chen, Weijie; Horsch, Karla; Lan, Li; Jamieson, Andrew R.; Sennett, Charlene A.; Jansen, Sanaz A.

In: Academic radiology, Vol. 15, No. 11, 01.11.2008, p. 1437-1445.

Research output: Contribution to journalArticle

Li, H, Giger, ML, Yuan, Y, Chen, W, Horsch, K, Lan, L, Jamieson, AR, Sennett, CA & Jansen, SA 2008, 'Evaluation of Computer-aided Diagnosis on a Large Clinical Full-field Digital Mammographic Dataset', Academic radiology, vol. 15, no. 11, pp. 1437-1445. https://doi.org/10.1016/j.acra.2008.05.004
Li, Hui ; Giger, Maryellen L. ; Yuan, Yading ; Chen, Weijie ; Horsch, Karla ; Lan, Li ; Jamieson, Andrew R. ; Sennett, Charlene A. ; Jansen, Sanaz A. / Evaluation of Computer-aided Diagnosis on a Large Clinical Full-field Digital Mammographic Dataset. In: Academic radiology. 2008 ; Vol. 15, No. 11. pp. 1437-1445.
@article{6b2d2036496a4390bb4e1050d76f12a3,
title = "Evaluation of Computer-aided Diagnosis on a Large Clinical Full-field Digital Mammographic Dataset",
abstract = "Rationale and Objectives: To convert and optimize our previously developed computerized analysis methods for use with images from full-field digital mammography (FFDM) for breast mass classification to aid in the diagnosis of breast cancer. Materials and Methods: An institutional review board approved protocol was obtained, with waiver of consent for retrospective use of mammograms and pathology data. Seven hundred thirty-nine FFDM images, which contained 287 biopsy-proven breast mass lesions, of which 148 lesions were malignant and 139 lesions were benign, were retrospectively collected. Lesion margins were delineated by an expert breast radiologist and were used as the truth for lesion-segmentation evaluation. Our computerized image analysis method consisted of several steps: 1) identified lesions were automatically extracted from the parenchymal background using computerized segmentation methods; 2) a set of image characteristics (mathematic descriptors) were automatically extracted from image data of the lesions and surrounding tissues; and 3) selected features were merged into an estimate of the probability of malignancy using a Bayesian artificial neural network classifier. Performance of the analyses was evaluated at various stages of the conversion using receiver-operating characteristic analysis. Results: An area under the curve value of 0.81 was obtained in the task of distinguishing between malignant and benign mass lesions in a round-robin by case evaluation on the entire FFDM dataset. We failed to show a statistically significant difference (P = .83) compared to results from our previous study in which the computerized classification was performed on digitized screen-film mammograms. Conclusions: Our computerized analysis methods developed on digitized screen-film mammography can be converted for use with FFDM. Results show that the computerized analysis methods for the diagnosis of breast mass lesions on FFDM are promising, and can potentially be used to aid clinicians in the diagnostic interpretation of FFDM.",
keywords = "breast mass classification, Computer-aided diagnosis, full-field digital mammography",
author = "Hui Li and Giger, {Maryellen L.} and Yading Yuan and Weijie Chen and Karla Horsch and Li Lan and Jamieson, {Andrew R.} and Sennett, {Charlene A.} and Jansen, {Sanaz A.}",
year = "2008",
month = "11",
day = "1",
doi = "10.1016/j.acra.2008.05.004",
language = "English (US)",
volume = "15",
pages = "1437--1445",
journal = "Academic Radiology",
issn = "1076-6332",
publisher = "Elsevier USA",
number = "11",

}

TY - JOUR

T1 - Evaluation of Computer-aided Diagnosis on a Large Clinical Full-field Digital Mammographic Dataset

AU - Li, Hui

AU - Giger, Maryellen L.

AU - Yuan, Yading

AU - Chen, Weijie

AU - Horsch, Karla

AU - Lan, Li

AU - Jamieson, Andrew R.

AU - Sennett, Charlene A.

AU - Jansen, Sanaz A.

PY - 2008/11/1

Y1 - 2008/11/1

N2 - Rationale and Objectives: To convert and optimize our previously developed computerized analysis methods for use with images from full-field digital mammography (FFDM) for breast mass classification to aid in the diagnosis of breast cancer. Materials and Methods: An institutional review board approved protocol was obtained, with waiver of consent for retrospective use of mammograms and pathology data. Seven hundred thirty-nine FFDM images, which contained 287 biopsy-proven breast mass lesions, of which 148 lesions were malignant and 139 lesions were benign, were retrospectively collected. Lesion margins were delineated by an expert breast radiologist and were used as the truth for lesion-segmentation evaluation. Our computerized image analysis method consisted of several steps: 1) identified lesions were automatically extracted from the parenchymal background using computerized segmentation methods; 2) a set of image characteristics (mathematic descriptors) were automatically extracted from image data of the lesions and surrounding tissues; and 3) selected features were merged into an estimate of the probability of malignancy using a Bayesian artificial neural network classifier. Performance of the analyses was evaluated at various stages of the conversion using receiver-operating characteristic analysis. Results: An area under the curve value of 0.81 was obtained in the task of distinguishing between malignant and benign mass lesions in a round-robin by case evaluation on the entire FFDM dataset. We failed to show a statistically significant difference (P = .83) compared to results from our previous study in which the computerized classification was performed on digitized screen-film mammograms. Conclusions: Our computerized analysis methods developed on digitized screen-film mammography can be converted for use with FFDM. Results show that the computerized analysis methods for the diagnosis of breast mass lesions on FFDM are promising, and can potentially be used to aid clinicians in the diagnostic interpretation of FFDM.

AB - Rationale and Objectives: To convert and optimize our previously developed computerized analysis methods for use with images from full-field digital mammography (FFDM) for breast mass classification to aid in the diagnosis of breast cancer. Materials and Methods: An institutional review board approved protocol was obtained, with waiver of consent for retrospective use of mammograms and pathology data. Seven hundred thirty-nine FFDM images, which contained 287 biopsy-proven breast mass lesions, of which 148 lesions were malignant and 139 lesions were benign, were retrospectively collected. Lesion margins were delineated by an expert breast radiologist and were used as the truth for lesion-segmentation evaluation. Our computerized image analysis method consisted of several steps: 1) identified lesions were automatically extracted from the parenchymal background using computerized segmentation methods; 2) a set of image characteristics (mathematic descriptors) were automatically extracted from image data of the lesions and surrounding tissues; and 3) selected features were merged into an estimate of the probability of malignancy using a Bayesian artificial neural network classifier. Performance of the analyses was evaluated at various stages of the conversion using receiver-operating characteristic analysis. Results: An area under the curve value of 0.81 was obtained in the task of distinguishing between malignant and benign mass lesions in a round-robin by case evaluation on the entire FFDM dataset. We failed to show a statistically significant difference (P = .83) compared to results from our previous study in which the computerized classification was performed on digitized screen-film mammograms. Conclusions: Our computerized analysis methods developed on digitized screen-film mammography can be converted for use with FFDM. Results show that the computerized analysis methods for the diagnosis of breast mass lesions on FFDM are promising, and can potentially be used to aid clinicians in the diagnostic interpretation of FFDM.

KW - breast mass classification

KW - Computer-aided diagnosis

KW - full-field digital mammography

UR - http://www.scopus.com/inward/record.url?scp=55349095801&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=55349095801&partnerID=8YFLogxK

U2 - 10.1016/j.acra.2008.05.004

DO - 10.1016/j.acra.2008.05.004

M3 - Article

C2 - 18995194

AN - SCOPUS:55349095801

VL - 15

SP - 1437

EP - 1445

JO - Academic Radiology

JF - Academic Radiology

SN - 1076-6332

IS - 11

ER -