Mutual information criterion for feature selection with application to classification of breast microcalcifications

Idit Diamant; Moran Shalhon; Jacob Goldberger; Hayit Greenspan

doi:10.1117/12.2216466

Mutual information criterion for feature selection with application to classification of breast microcalcifications

Idit Diamant, Moran Shalhon, Jacob Goldberger, Hayit Greenspan

Bar-Ilan University - The Alexander Kofkin Faculty of Engineering

Tel Aviv University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Scopus citations

Abstract

Classification of clustered breast microcalcifications into benign and malignant categories is an extremely challenging task for computerized algorithms and expert radiologists alike. In this paper we present a novel method for feature selection based on mutual information (MI) criterion for automatic classification of microcalcifications. We explored the MI based feature selection for various texture features. The proposed method was evaluated on a standardized digital database for screening mammography (DDSM). Experimental results demonstrate the effectiveness and the advantage of using the MI-based feature selection to obtain the most relevant features for the task and thus to provide for improved performance as compared to using all features.

Original language	English
Title of host publication	Medical Imaging 2016
Subtitle of host publication	Image Processing
Editors	Martin A. Styner, Elsa D. Angelini, Elsa D. Angelini
Publisher	SPIE
ISBN (Electronic)	9781510600195
DOIs	https://doi.org/10.1117/12.2216466
State	Published - 2016
Event	Medical Imaging 2016: Image Processing - San Diego, United States Duration: 1 Mar 2016 → 3 Mar 2016

Publication series

Name	Progress in Biomedical Optics and Imaging - Proceedings of SPIE
Volume	9784
ISSN (Print)	1605-7422

Conference

Conference	Medical Imaging 2016: Image Processing
Country/Territory	United States
City	San Diego
Period	1/03/16 → 3/03/16

Bibliographical note

Publisher Copyright:
© 2016 SPIE.

Keywords

Computer-aided diagnosis (CADx)
Feature selection
Mammography
Microcalcifications
Mutual information
Texture features

Access to Document

10.1117/12.2216466

Cite this

Diamant, I., Shalhon, M., Goldberger, J., & Greenspan, H. (2016). Mutual information criterion for feature selection with application to classification of breast microcalcifications. In M. A. Styner, E. D. Angelini, & E. D. Angelini (Eds.), Medical Imaging 2016: Image Processing Article 97841S (Progress in Biomedical Optics and Imaging - Proceedings of SPIE; Vol. 9784). SPIE. https://doi.org/10.1117/12.2216466

@inproceedings{4c7d7f2194ff4730b245716ee3db2a1a,

title = "Mutual information criterion for feature selection with application to classification of breast microcalcifications",

abstract = "Classification of clustered breast microcalcifications into benign and malignant categories is an extremely challenging task for computerized algorithms and expert radiologists alike. In this paper we present a novel method for feature selection based on mutual information (MI) criterion for automatic classification of microcalcifications. We explored the MI based feature selection for various texture features. The proposed method was evaluated on a standardized digital database for screening mammography (DDSM). Experimental results demonstrate the effectiveness and the advantage of using the MI-based feature selection to obtain the most relevant features for the task and thus to provide for improved performance as compared to using all features.",

keywords = "Computer-aided diagnosis (CADx), Feature selection, Mammography, Microcalcifications, Mutual information, Texture features",

author = "Idit Diamant and Moran Shalhon and Jacob Goldberger and Hayit Greenspan",

note = "Publisher Copyright: {\textcopyright} 2016 SPIE.; Medical Imaging 2016: Image Processing ; Conference date: 01-03-2016 Through 03-03-2016",

year = "2016",

doi = "10.1117/12.2216466",

language = "אנגלית",

series = "Progress in Biomedical Optics and Imaging - Proceedings of SPIE",

publisher = "SPIE",

editor = "Styner, {Martin A.} and Angelini, {Elsa D.} and Angelini, {Elsa D.}",

booktitle = "Medical Imaging 2016",

address = "ארצות הברית",

}

Diamant, I, Shalhon, M, Goldberger, J & Greenspan, H 2016, Mutual information criterion for feature selection with application to classification of breast microcalcifications. in MA Styner, ED Angelini & ED Angelini (eds), Medical Imaging 2016: Image Processing., 97841S, Progress in Biomedical Optics and Imaging - Proceedings of SPIE, vol. 9784, SPIE, Medical Imaging 2016: Image Processing, San Diego, United States, 1/03/16. https://doi.org/10.1117/12.2216466

Mutual information criterion for feature selection with application to classification of breast microcalcifications. / Diamant, Idit; Shalhon, Moran; Goldberger, Jacob et al.
Medical Imaging 2016: Image Processing. ed. / Martin A. Styner; Elsa D. Angelini; Elsa D. Angelini. SPIE, 2016. 97841S (Progress in Biomedical Optics and Imaging - Proceedings of SPIE; Vol. 9784).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Mutual information criterion for feature selection with application to classification of breast microcalcifications

AU - Diamant, Idit

AU - Shalhon, Moran

AU - Goldberger, Jacob

AU - Greenspan, Hayit

PY - 2016

Y1 - 2016

N2 - Classification of clustered breast microcalcifications into benign and malignant categories is an extremely challenging task for computerized algorithms and expert radiologists alike. In this paper we present a novel method for feature selection based on mutual information (MI) criterion for automatic classification of microcalcifications. We explored the MI based feature selection for various texture features. The proposed method was evaluated on a standardized digital database for screening mammography (DDSM). Experimental results demonstrate the effectiveness and the advantage of using the MI-based feature selection to obtain the most relevant features for the task and thus to provide for improved performance as compared to using all features.

AB - Classification of clustered breast microcalcifications into benign and malignant categories is an extremely challenging task for computerized algorithms and expert radiologists alike. In this paper we present a novel method for feature selection based on mutual information (MI) criterion for automatic classification of microcalcifications. We explored the MI based feature selection for various texture features. The proposed method was evaluated on a standardized digital database for screening mammography (DDSM). Experimental results demonstrate the effectiveness and the advantage of using the MI-based feature selection to obtain the most relevant features for the task and thus to provide for improved performance as compared to using all features.

KW - Computer-aided diagnosis (CADx)

KW - Feature selection

KW - Mammography

KW - Microcalcifications

KW - Mutual information

KW - Texture features

UR - http://www.scopus.com/inward/record.url?scp=84981722767&partnerID=8YFLogxK

U2 - 10.1117/12.2216466

DO - 10.1117/12.2216466

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:84981722767

T3 - Progress in Biomedical Optics and Imaging - Proceedings of SPIE

BT - Medical Imaging 2016

A2 - Styner, Martin A.

A2 - Angelini, Elsa D.

PB - SPIE

T2 - Medical Imaging 2016: Image Processing

Y2 - 1 March 2016 through 3 March 2016

ER -

Mutual information criterion for feature selection with application to classification of breast microcalcifications

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this