Multiobjective genetic optimization of diagnostic classifiers used in the computerized detection of mass lesions in mammography

Matthew A Kupinski, Mark A. Anastasio, Maryellen L. Giger

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

4 Citations (Scopus)

Abstract

We have recently proposed and developed a multiobjective approach to training classification systems. In this approach, the objectives, i.e., the sensitivity and specificity, of a classifier are simultaneously optimized, resulting in a series of solutions that are equivalent in the absence of any a priori knowledge regarding the relative merits of the two objectives. These solutions form a receiver operating characteristic (ROC) curve that is, theoretically, the best possible ROC curve that can be obtained using the given classifier and given training dataset. We have applied this technique to the optimization of classifiers for the computerized detection of mass lesions in digitized mammograms. Comparisons will be made between the results obtained using the multiobjective approach and results obtained using more conventional approaches. We employed a database of 60 consecutive, non-palpable mass lesion cases. Features relating to the geometry, intensity, and gradients of the images were calculated for each visible lesion and for many false detections. Using a conventionally trained linear classifier we were able to achieve an Az of 0.84 while the multiobjective approach to training a linear classifier yielded an Az of 0.87 in the task of distinguishing between true lesions and false detections. Using a multiobjective approach to train a rule-based classifier with 5 thresholding rules resulted in an Az of 0.88 in the task of distinguishing between true lesions and false detections.

Original languageEnglish (US)
Title of host publicationProceedings of SPIE - The International Society for Optical Engineering
PublisherSociety of Photo-Optical Instrumentation Engineers
Volume3979
StatePublished - 2000
Externally publishedYes
EventMedical Imaging 2000: Image Processing - San Diego, CA, USA
Duration: Feb 14 2000Feb 17 2000

Other

OtherMedical Imaging 2000: Image Processing
CitySan Diego, CA, USA
Period2/14/002/17/00

Fingerprint

Mammography
classifiers
lesions
Classifiers
optimization
education
receivers
curves
gradients
Geometry
sensitivity
geometry

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Condensed Matter Physics

Cite this

Kupinski, M. A., Anastasio, M. A., & Giger, M. L. (2000). Multiobjective genetic optimization of diagnostic classifiers used in the computerized detection of mass lesions in mammography. In Proceedings of SPIE - The International Society for Optical Engineering (Vol. 3979). Society of Photo-Optical Instrumentation Engineers.

Multiobjective genetic optimization of diagnostic classifiers used in the computerized detection of mass lesions in mammography. / Kupinski, Matthew A; Anastasio, Mark A.; Giger, Maryellen L.

Proceedings of SPIE - The International Society for Optical Engineering. Vol. 3979 Society of Photo-Optical Instrumentation Engineers, 2000.

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

Kupinski, MA, Anastasio, MA & Giger, ML 2000, Multiobjective genetic optimization of diagnostic classifiers used in the computerized detection of mass lesions in mammography. in Proceedings of SPIE - The International Society for Optical Engineering. vol. 3979, Society of Photo-Optical Instrumentation Engineers, Medical Imaging 2000: Image Processing, San Diego, CA, USA, 2/14/00.
Kupinski MA, Anastasio MA, Giger ML. Multiobjective genetic optimization of diagnostic classifiers used in the computerized detection of mass lesions in mammography. In Proceedings of SPIE - The International Society for Optical Engineering. Vol. 3979. Society of Photo-Optical Instrumentation Engineers. 2000
Kupinski, Matthew A ; Anastasio, Mark A. ; Giger, Maryellen L. / Multiobjective genetic optimization of diagnostic classifiers used in the computerized detection of mass lesions in mammography. Proceedings of SPIE - The International Society for Optical Engineering. Vol. 3979 Society of Photo-Optical Instrumentation Engineers, 2000.
@inproceedings{c22f5bf17e1a499a95c1f96e32342200,
title = "Multiobjective genetic optimization of diagnostic classifiers used in the computerized detection of mass lesions in mammography",
abstract = "We have recently proposed and developed a multiobjective approach to training classification systems. In this approach, the objectives, i.e., the sensitivity and specificity, of a classifier are simultaneously optimized, resulting in a series of solutions that are equivalent in the absence of any a priori knowledge regarding the relative merits of the two objectives. These solutions form a receiver operating characteristic (ROC) curve that is, theoretically, the best possible ROC curve that can be obtained using the given classifier and given training dataset. We have applied this technique to the optimization of classifiers for the computerized detection of mass lesions in digitized mammograms. Comparisons will be made between the results obtained using the multiobjective approach and results obtained using more conventional approaches. We employed a database of 60 consecutive, non-palpable mass lesion cases. Features relating to the geometry, intensity, and gradients of the images were calculated for each visible lesion and for many false detections. Using a conventionally trained linear classifier we were able to achieve an Az of 0.84 while the multiobjective approach to training a linear classifier yielded an Az of 0.87 in the task of distinguishing between true lesions and false detections. Using a multiobjective approach to train a rule-based classifier with 5 thresholding rules resulted in an Az of 0.88 in the task of distinguishing between true lesions and false detections.",
author = "Kupinski, {Matthew A} and Anastasio, {Mark A.} and Giger, {Maryellen L.}",
year = "2000",
language = "English (US)",
volume = "3979",
booktitle = "Proceedings of SPIE - The International Society for Optical Engineering",
publisher = "Society of Photo-Optical Instrumentation Engineers",

}

TY - GEN

T1 - Multiobjective genetic optimization of diagnostic classifiers used in the computerized detection of mass lesions in mammography

AU - Kupinski, Matthew A

AU - Anastasio, Mark A.

AU - Giger, Maryellen L.

PY - 2000

Y1 - 2000

N2 - We have recently proposed and developed a multiobjective approach to training classification systems. In this approach, the objectives, i.e., the sensitivity and specificity, of a classifier are simultaneously optimized, resulting in a series of solutions that are equivalent in the absence of any a priori knowledge regarding the relative merits of the two objectives. These solutions form a receiver operating characteristic (ROC) curve that is, theoretically, the best possible ROC curve that can be obtained using the given classifier and given training dataset. We have applied this technique to the optimization of classifiers for the computerized detection of mass lesions in digitized mammograms. Comparisons will be made between the results obtained using the multiobjective approach and results obtained using more conventional approaches. We employed a database of 60 consecutive, non-palpable mass lesion cases. Features relating to the geometry, intensity, and gradients of the images were calculated for each visible lesion and for many false detections. Using a conventionally trained linear classifier we were able to achieve an Az of 0.84 while the multiobjective approach to training a linear classifier yielded an Az of 0.87 in the task of distinguishing between true lesions and false detections. Using a multiobjective approach to train a rule-based classifier with 5 thresholding rules resulted in an Az of 0.88 in the task of distinguishing between true lesions and false detections.

AB - We have recently proposed and developed a multiobjective approach to training classification systems. In this approach, the objectives, i.e., the sensitivity and specificity, of a classifier are simultaneously optimized, resulting in a series of solutions that are equivalent in the absence of any a priori knowledge regarding the relative merits of the two objectives. These solutions form a receiver operating characteristic (ROC) curve that is, theoretically, the best possible ROC curve that can be obtained using the given classifier and given training dataset. We have applied this technique to the optimization of classifiers for the computerized detection of mass lesions in digitized mammograms. Comparisons will be made between the results obtained using the multiobjective approach and results obtained using more conventional approaches. We employed a database of 60 consecutive, non-palpable mass lesion cases. Features relating to the geometry, intensity, and gradients of the images were calculated for each visible lesion and for many false detections. Using a conventionally trained linear classifier we were able to achieve an Az of 0.84 while the multiobjective approach to training a linear classifier yielded an Az of 0.87 in the task of distinguishing between true lesions and false detections. Using a multiobjective approach to train a rule-based classifier with 5 thresholding rules resulted in an Az of 0.88 in the task of distinguishing between true lesions and false detections.

UR - http://www.scopus.com/inward/record.url?scp=0033686017&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033686017&partnerID=8YFLogxK

M3 - Conference contribution

VL - 3979

BT - Proceedings of SPIE - The International Society for Optical Engineering

PB - Society of Photo-Optical Instrumentation Engineers

ER -