The reliability issue of computer-aided breast cancer diagnosis

Boris Kovalerchuk, Evangelos Triantaphyllou, James F. Ruiz, Vetle I. Torvik, Evgeni Vityaev

Research output: Contribution to journalArticle

Abstract

This paper introduces a number of reliability criteria for computer-aided diagnostic systems for breast cancer. These criteria are then used to analyze some published neural network systems. It is also shown that the property of monotonicity for the data is rather natural in this medical domain, and it has the potential to significantly improve the reliability of breast cancer diagnosis while maintaining a general representation power. A central part of this paper is devoted to the representation/narrow vicinity hypothesis, upon which existing computer-aided diagnostic methods heavily rely. The paper also develops a framework for determining the validity of this hypothesis. The same framework can be used to construct a diagnostic procedure with improved reliability. (C) 2000 Academic Press.

Original languageEnglish (US)
Pages (from-to)296-313
Number of pages18
JournalComputers and Biomedical Research
Volume33
Issue number4
DOIs
StatePublished - Jan 1 2000
Externally publishedYes

Fingerprint

Breast Neoplasms
Neural networks

Keywords

  • Computer-aided diagnostic systems
  • Data monotonicity
  • Discriminant analysis
  • Machine learning
  • Neural networks
  • Reliability of diagnosis
  • Representation/narrow vicinity hypothesis

ASJC Scopus subject areas

  • Medicine (miscellaneous)

Cite this

The reliability issue of computer-aided breast cancer diagnosis. / Kovalerchuk, Boris; Triantaphyllou, Evangelos; Ruiz, James F.; Torvik, Vetle I.; Vityaev, Evgeni.

In: Computers and Biomedical Research, Vol. 33, No. 4, 01.01.2000, p. 296-313.

Research output: Contribution to journalArticle

Kovalerchuk, B, Triantaphyllou, E, Ruiz, JF, Torvik, VI & Vityaev, E 2000, 'The reliability issue of computer-aided breast cancer diagnosis', Computers and Biomedical Research, vol. 33, no. 4, pp. 296-313. https://doi.org/10.1006/cbmr.2000.1546
Kovalerchuk, Boris ; Triantaphyllou, Evangelos ; Ruiz, James F. ; Torvik, Vetle I. ; Vityaev, Evgeni. / The reliability issue of computer-aided breast cancer diagnosis. In: Computers and Biomedical Research. 2000 ; Vol. 33, No. 4. pp. 296-313.
@article{a2cbb4b820d3407a82c9052616485c07,
title = "The reliability issue of computer-aided breast cancer diagnosis",
abstract = "This paper introduces a number of reliability criteria for computer-aided diagnostic systems for breast cancer. These criteria are then used to analyze some published neural network systems. It is also shown that the property of monotonicity for the data is rather natural in this medical domain, and it has the potential to significantly improve the reliability of breast cancer diagnosis while maintaining a general representation power. A central part of this paper is devoted to the representation/narrow vicinity hypothesis, upon which existing computer-aided diagnostic methods heavily rely. The paper also develops a framework for determining the validity of this hypothesis. The same framework can be used to construct a diagnostic procedure with improved reliability. (C) 2000 Academic Press.",
keywords = "Computer-aided diagnostic systems, Data monotonicity, Discriminant analysis, Machine learning, Neural networks, Reliability of diagnosis, Representation/narrow vicinity hypothesis",
author = "Boris Kovalerchuk and Evangelos Triantaphyllou and Ruiz, {James F.} and Torvik, {Vetle I.} and Evgeni Vityaev",
year = "2000",
month = "1",
day = "1",
doi = "10.1006/cbmr.2000.1546",
language = "English (US)",
volume = "33",
pages = "296--313",
journal = "Journal of Biomedical Informatics",
issn = "1532-0464",
publisher = "Academic Press Inc.",
number = "4",

}

TY - JOUR

T1 - The reliability issue of computer-aided breast cancer diagnosis

AU - Kovalerchuk, Boris

AU - Triantaphyllou, Evangelos

AU - Ruiz, James F.

AU - Torvik, Vetle I.

AU - Vityaev, Evgeni

PY - 2000/1/1

Y1 - 2000/1/1

N2 - This paper introduces a number of reliability criteria for computer-aided diagnostic systems for breast cancer. These criteria are then used to analyze some published neural network systems. It is also shown that the property of monotonicity for the data is rather natural in this medical domain, and it has the potential to significantly improve the reliability of breast cancer diagnosis while maintaining a general representation power. A central part of this paper is devoted to the representation/narrow vicinity hypothesis, upon which existing computer-aided diagnostic methods heavily rely. The paper also develops a framework for determining the validity of this hypothesis. The same framework can be used to construct a diagnostic procedure with improved reliability. (C) 2000 Academic Press.

AB - This paper introduces a number of reliability criteria for computer-aided diagnostic systems for breast cancer. These criteria are then used to analyze some published neural network systems. It is also shown that the property of monotonicity for the data is rather natural in this medical domain, and it has the potential to significantly improve the reliability of breast cancer diagnosis while maintaining a general representation power. A central part of this paper is devoted to the representation/narrow vicinity hypothesis, upon which existing computer-aided diagnostic methods heavily rely. The paper also develops a framework for determining the validity of this hypothesis. The same framework can be used to construct a diagnostic procedure with improved reliability. (C) 2000 Academic Press.

KW - Computer-aided diagnostic systems

KW - Data monotonicity

KW - Discriminant analysis

KW - Machine learning

KW - Neural networks

KW - Reliability of diagnosis

KW - Representation/narrow vicinity hypothesis

UR - http://www.scopus.com/inward/record.url?scp=0033843239&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033843239&partnerID=8YFLogxK

U2 - 10.1006/cbmr.2000.1546

DO - 10.1006/cbmr.2000.1546

M3 - Article

C2 - 10944406

AN - SCOPUS:0033843239

VL - 33

SP - 296

EP - 313

JO - Journal of Biomedical Informatics

JF - Journal of Biomedical Informatics

SN - 1532-0464

IS - 4

ER -