A multiple instance learning approach toward optimal classification of pathology slides

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

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

23 Citations (Scopus)

Abstract

Pathology slides are diagnosed based on the histological descriptors extracted from regions of interest (ROIs) identified on each slide by the pathologists. A slide usually contains multiple regions of interest and a positive (cancer) diagnosis is confirmed when at least one of the ROIs in the slide is identified as positive. For a negative diagnosis the pathologist has to rule out cancer for each and every ROI available. Our research is motivated toward computer-assisted classification of digitized slides. The objective in this study is to develop a classifier to optimize classification accuracy at the slide level. Traditional supervised training techniques which are trained to optimize classifier performance at the ROI level yield suboptimal performance in this problem. We propose a multiple instance learning approach based on the implementation of the large margin principle with different loss functions defined for positive and negative samples. We consider the classification of intraductal breast lesions as a case study, and perform experimental studies comparing our approach against the state-of-the-art.

Original languageEnglish
Title of host publicationProceedings - International Conference on Pattern Recognition
Pages2732-2735
Number of pages4
DOIs
StatePublished - 2010
Event2010 20th International Conference on Pattern Recognition, ICPR 2010 - Istanbul, Turkey
Duration: Aug 23 2010Aug 26 2010

Other

Other2010 20th International Conference on Pattern Recognition, ICPR 2010
CountryTurkey
CityIstanbul
Period8/23/108/26/10

Fingerprint

Pathology
Classifiers

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition

Cite this

Dundar, M. M., Badve, S., Raykar, V. C., Jain, R. K., Sertel, O., & Gurcan, M. N. (2010). A multiple instance learning approach toward optimal classification of pathology slides. In Proceedings - International Conference on Pattern Recognition (pp. 2732-2735). [5596023] https://doi.org/10.1109/ICPR.2010.669

A multiple instance learning approach toward optimal classification of pathology slides. / Dundar, M. Murat; Badve, Sunil; Raykar, Vikas C.; Jain, Rohit K.; Sertel, Olcay; Gurcan, Metin N.

Proceedings - International Conference on Pattern Recognition. 2010. p. 2732-2735 5596023.

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

Dundar, MM, Badve, S, Raykar, VC, Jain, RK, Sertel, O & Gurcan, MN 2010, A multiple instance learning approach toward optimal classification of pathology slides. in Proceedings - International Conference on Pattern Recognition., 5596023, pp. 2732-2735, 2010 20th International Conference on Pattern Recognition, ICPR 2010, Istanbul, Turkey, 8/23/10. https://doi.org/10.1109/ICPR.2010.669
Dundar MM, Badve S, Raykar VC, Jain RK, Sertel O, Gurcan MN. A multiple instance learning approach toward optimal classification of pathology slides. In Proceedings - International Conference on Pattern Recognition. 2010. p. 2732-2735. 5596023 https://doi.org/10.1109/ICPR.2010.669
Dundar, M. Murat ; Badve, Sunil ; Raykar, Vikas C. ; Jain, Rohit K. ; Sertel, Olcay ; Gurcan, Metin N. / A multiple instance learning approach toward optimal classification of pathology slides. Proceedings - International Conference on Pattern Recognition. 2010. pp. 2732-2735
@inproceedings{bf96d917c4534bd89bf1b1216892d7b7,
title = "A multiple instance learning approach toward optimal classification of pathology slides",
abstract = "Pathology slides are diagnosed based on the histological descriptors extracted from regions of interest (ROIs) identified on each slide by the pathologists. A slide usually contains multiple regions of interest and a positive (cancer) diagnosis is confirmed when at least one of the ROIs in the slide is identified as positive. For a negative diagnosis the pathologist has to rule out cancer for each and every ROI available. Our research is motivated toward computer-assisted classification of digitized slides. The objective in this study is to develop a classifier to optimize classification accuracy at the slide level. Traditional supervised training techniques which are trained to optimize classifier performance at the ROI level yield suboptimal performance in this problem. We propose a multiple instance learning approach based on the implementation of the large margin principle with different loss functions defined for positive and negative samples. We consider the classification of intraductal breast lesions as a case study, and perform experimental studies comparing our approach against the state-of-the-art.",
author = "Dundar, {M. Murat} and Sunil Badve and Raykar, {Vikas C.} and Jain, {Rohit K.} and Olcay Sertel and Gurcan, {Metin N.}",
year = "2010",
doi = "10.1109/ICPR.2010.669",
language = "English",
isbn = "9780769541099",
pages = "2732--2735",
booktitle = "Proceedings - International Conference on Pattern Recognition",

}

TY - GEN

T1 - A multiple instance learning approach toward optimal classification of pathology slides

AU - Dundar, M. Murat

AU - Badve, Sunil

AU - Raykar, Vikas C.

AU - Jain, Rohit K.

AU - Sertel, Olcay

AU - Gurcan, Metin N.

PY - 2010

Y1 - 2010

N2 - Pathology slides are diagnosed based on the histological descriptors extracted from regions of interest (ROIs) identified on each slide by the pathologists. A slide usually contains multiple regions of interest and a positive (cancer) diagnosis is confirmed when at least one of the ROIs in the slide is identified as positive. For a negative diagnosis the pathologist has to rule out cancer for each and every ROI available. Our research is motivated toward computer-assisted classification of digitized slides. The objective in this study is to develop a classifier to optimize classification accuracy at the slide level. Traditional supervised training techniques which are trained to optimize classifier performance at the ROI level yield suboptimal performance in this problem. We propose a multiple instance learning approach based on the implementation of the large margin principle with different loss functions defined for positive and negative samples. We consider the classification of intraductal breast lesions as a case study, and perform experimental studies comparing our approach against the state-of-the-art.

AB - Pathology slides are diagnosed based on the histological descriptors extracted from regions of interest (ROIs) identified on each slide by the pathologists. A slide usually contains multiple regions of interest and a positive (cancer) diagnosis is confirmed when at least one of the ROIs in the slide is identified as positive. For a negative diagnosis the pathologist has to rule out cancer for each and every ROI available. Our research is motivated toward computer-assisted classification of digitized slides. The objective in this study is to develop a classifier to optimize classification accuracy at the slide level. Traditional supervised training techniques which are trained to optimize classifier performance at the ROI level yield suboptimal performance in this problem. We propose a multiple instance learning approach based on the implementation of the large margin principle with different loss functions defined for positive and negative samples. We consider the classification of intraductal breast lesions as a case study, and perform experimental studies comparing our approach against the state-of-the-art.

UR - http://www.scopus.com/inward/record.url?scp=78149479937&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78149479937&partnerID=8YFLogxK

U2 - 10.1109/ICPR.2010.669

DO - 10.1109/ICPR.2010.669

M3 - Conference contribution

AN - SCOPUS:78149479937

SN - 9780769541099

SP - 2732

EP - 2735

BT - Proceedings - International Conference on Pattern Recognition

ER -