Computerized classification of intraductal breast lesions using histopathological images

M. Murat Dundar, Sunil Badve, Gokhan Bilgin, Vikas Raykar, Rohit Jain, Olcay Sertel, Metin N. Gurcan

Research output: Contribution to journalArticle

88 Citations (Scopus)

Abstract

In the diagnosis of preinvasive breast cancer, some of the intraductal proliferations pose a special challenge. The continuum of intraductal breast lesions includes the usual ductal hyperplasia (UDH), atypical ductal hyperplasia (ADH), and ductal carcinoma in situ (DCIS). The current standard of care is to perform percutaneous needle biopsies for diagnosis of palpable and image-detected breast abnormalities. UDH is considered benign and patients diagnosed UDH undergo routine follow-up, whereas ADH and DCIS are considered actionable and patients diagnosed with these two subtypes get additional surgical procedures. About 250000 new cases of intraductal breast lesions are diagnosed every year. A conservative estimate would suggest that at least 50% of these patients are needlessly undergoing unnecessary surgeries. Thus, improvement in the diagnostic reproducibility and accuracy is critically important for effective clinical management of these patients. In this study, a prototype system for automatically classifying breast microscopic tissues to distinguish between UDH and actionable subtypes (ADH and DCIS) is introduced. This system automatically evaluates digitized slides of tissues for certain cytological criteria and classifies the tissues based on the quantitative features derived from the images. The system is trained using a total of 327 regions of interest (ROIs) collected across 62 patient cases and tested with a sequestered set of 149 ROIs collected across 33 patient cases. An overall accuracy of 87.9% is achieved on the entire test data. The test accuracy of 84.6% is obtained with borderline cases (26 of the 33 test cases) only, when compared against the diagnostic accuracies of nine pathologists on the same set (81.2% average), indicates that the system is highly competitive with the expert pathologists as a stand-alone diagnostic tool and has a great potential in improving diagnostic accuracy and reproducibility when used as a second reader in conjunction with the pathologists.

Original languageEnglish
Article number5706360
Pages (from-to)1977-1984
Number of pages8
JournalIEEE Transactions on Biomedical Engineering
Volume58
Issue number7
DOIs
StatePublished - Jul 2011

Fingerprint

Tissue
Biopsy
Needles
Surgery

Keywords

  • Cell segmentation
  • computer-aided diagnosis
  • histopathological image analysis
  • intraductal breast lesions
  • multiple instance learning

ASJC Scopus subject areas

  • Biomedical Engineering

Cite this

Computerized classification of intraductal breast lesions using histopathological images. / Dundar, M. Murat; Badve, Sunil; Bilgin, Gokhan; Raykar, Vikas; Jain, Rohit; Sertel, Olcay; Gurcan, Metin N.

In: IEEE Transactions on Biomedical Engineering, Vol. 58, No. 7, 5706360, 07.2011, p. 1977-1984.

Research output: Contribution to journalArticle

Dundar, M. Murat ; Badve, Sunil ; Bilgin, Gokhan ; Raykar, Vikas ; Jain, Rohit ; Sertel, Olcay ; Gurcan, Metin N. / Computerized classification of intraductal breast lesions using histopathological images. In: IEEE Transactions on Biomedical Engineering. 2011 ; Vol. 58, No. 7. pp. 1977-1984.
@article{a22ba20687a447fb85b103db64fb04ca,
title = "Computerized classification of intraductal breast lesions using histopathological images",
abstract = "In the diagnosis of preinvasive breast cancer, some of the intraductal proliferations pose a special challenge. The continuum of intraductal breast lesions includes the usual ductal hyperplasia (UDH), atypical ductal hyperplasia (ADH), and ductal carcinoma in situ (DCIS). The current standard of care is to perform percutaneous needle biopsies for diagnosis of palpable and image-detected breast abnormalities. UDH is considered benign and patients diagnosed UDH undergo routine follow-up, whereas ADH and DCIS are considered actionable and patients diagnosed with these two subtypes get additional surgical procedures. About 250000 new cases of intraductal breast lesions are diagnosed every year. A conservative estimate would suggest that at least 50{\%} of these patients are needlessly undergoing unnecessary surgeries. Thus, improvement in the diagnostic reproducibility and accuracy is critically important for effective clinical management of these patients. In this study, a prototype system for automatically classifying breast microscopic tissues to distinguish between UDH and actionable subtypes (ADH and DCIS) is introduced. This system automatically evaluates digitized slides of tissues for certain cytological criteria and classifies the tissues based on the quantitative features derived from the images. The system is trained using a total of 327 regions of interest (ROIs) collected across 62 patient cases and tested with a sequestered set of 149 ROIs collected across 33 patient cases. An overall accuracy of 87.9{\%} is achieved on the entire test data. The test accuracy of 84.6{\%} is obtained with borderline cases (26 of the 33 test cases) only, when compared against the diagnostic accuracies of nine pathologists on the same set (81.2{\%} average), indicates that the system is highly competitive with the expert pathologists as a stand-alone diagnostic tool and has a great potential in improving diagnostic accuracy and reproducibility when used as a second reader in conjunction with the pathologists.",
keywords = "Cell segmentation, computer-aided diagnosis, histopathological image analysis, intraductal breast lesions, multiple instance learning",
author = "Dundar, {M. Murat} and Sunil Badve and Gokhan Bilgin and Vikas Raykar and Rohit Jain and Olcay Sertel and Gurcan, {Metin N.}",
year = "2011",
month = "7",
doi = "10.1109/TBME.2011.2110648",
language = "English",
volume = "58",
pages = "1977--1984",
journal = "IEEE Transactions on Biomedical Engineering",
issn = "0018-9294",
publisher = "IEEE Computer Society",
number = "7",

}

TY - JOUR

T1 - Computerized classification of intraductal breast lesions using histopathological images

AU - Dundar, M. Murat

AU - Badve, Sunil

AU - Bilgin, Gokhan

AU - Raykar, Vikas

AU - Jain, Rohit

AU - Sertel, Olcay

AU - Gurcan, Metin N.

PY - 2011/7

Y1 - 2011/7

N2 - In the diagnosis of preinvasive breast cancer, some of the intraductal proliferations pose a special challenge. The continuum of intraductal breast lesions includes the usual ductal hyperplasia (UDH), atypical ductal hyperplasia (ADH), and ductal carcinoma in situ (DCIS). The current standard of care is to perform percutaneous needle biopsies for diagnosis of palpable and image-detected breast abnormalities. UDH is considered benign and patients diagnosed UDH undergo routine follow-up, whereas ADH and DCIS are considered actionable and patients diagnosed with these two subtypes get additional surgical procedures. About 250000 new cases of intraductal breast lesions are diagnosed every year. A conservative estimate would suggest that at least 50% of these patients are needlessly undergoing unnecessary surgeries. Thus, improvement in the diagnostic reproducibility and accuracy is critically important for effective clinical management of these patients. In this study, a prototype system for automatically classifying breast microscopic tissues to distinguish between UDH and actionable subtypes (ADH and DCIS) is introduced. This system automatically evaluates digitized slides of tissues for certain cytological criteria and classifies the tissues based on the quantitative features derived from the images. The system is trained using a total of 327 regions of interest (ROIs) collected across 62 patient cases and tested with a sequestered set of 149 ROIs collected across 33 patient cases. An overall accuracy of 87.9% is achieved on the entire test data. The test accuracy of 84.6% is obtained with borderline cases (26 of the 33 test cases) only, when compared against the diagnostic accuracies of nine pathologists on the same set (81.2% average), indicates that the system is highly competitive with the expert pathologists as a stand-alone diagnostic tool and has a great potential in improving diagnostic accuracy and reproducibility when used as a second reader in conjunction with the pathologists.

AB - In the diagnosis of preinvasive breast cancer, some of the intraductal proliferations pose a special challenge. The continuum of intraductal breast lesions includes the usual ductal hyperplasia (UDH), atypical ductal hyperplasia (ADH), and ductal carcinoma in situ (DCIS). The current standard of care is to perform percutaneous needle biopsies for diagnosis of palpable and image-detected breast abnormalities. UDH is considered benign and patients diagnosed UDH undergo routine follow-up, whereas ADH and DCIS are considered actionable and patients diagnosed with these two subtypes get additional surgical procedures. About 250000 new cases of intraductal breast lesions are diagnosed every year. A conservative estimate would suggest that at least 50% of these patients are needlessly undergoing unnecessary surgeries. Thus, improvement in the diagnostic reproducibility and accuracy is critically important for effective clinical management of these patients. In this study, a prototype system for automatically classifying breast microscopic tissues to distinguish between UDH and actionable subtypes (ADH and DCIS) is introduced. This system automatically evaluates digitized slides of tissues for certain cytological criteria and classifies the tissues based on the quantitative features derived from the images. The system is trained using a total of 327 regions of interest (ROIs) collected across 62 patient cases and tested with a sequestered set of 149 ROIs collected across 33 patient cases. An overall accuracy of 87.9% is achieved on the entire test data. The test accuracy of 84.6% is obtained with borderline cases (26 of the 33 test cases) only, when compared against the diagnostic accuracies of nine pathologists on the same set (81.2% average), indicates that the system is highly competitive with the expert pathologists as a stand-alone diagnostic tool and has a great potential in improving diagnostic accuracy and reproducibility when used as a second reader in conjunction with the pathologists.

KW - Cell segmentation

KW - computer-aided diagnosis

KW - histopathological image analysis

KW - intraductal breast lesions

KW - multiple instance learning

UR - http://www.scopus.com/inward/record.url?scp=79959565637&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79959565637&partnerID=8YFLogxK

U2 - 10.1109/TBME.2011.2110648

DO - 10.1109/TBME.2011.2110648

M3 - Article

VL - 58

SP - 1977

EP - 1984

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

SN - 0018-9294

IS - 7

M1 - 5706360

ER -