An integrated index for breast cancer identification using histogram of oriented gradient and kernel locality preserving projection features extracted from thermograms

U. Raghavendra, U. Rajendra Acharya, E. Y.K. Ng, Jen Hong Tan, Anjan Gudigar

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Breast cancer is one of the prime causes of death in women worldwide. Thermography has shown a great potential in screening the breast cancer and overcomes the limitations of mammography. Moreover, interpretations of thermogram images are dependent on the specialists, which may lead to errors and uneven results. Preliminary screening method should detect the hazardous, destructive tumours effectively to improve the accuracy. The growth of malignant tumour can increase the internal temperature which can be captured by thermograms. Thus in this work, locally normalised histogram of oriented gradients (HOG) based preliminary screening computer aided diagnosis tool is proposed. HOG is able to record the minute internal variations in thermograms. In order to reduce the dimensions of extracted HOG descriptors kernel locality preserving projection (KLPP) is used. The resulting KLPP features are then ranked to form an efficient classification model. Various machine learning algorithms are used to validate the proposed method. Our method shows a promising performance with an average accuracy, sensitivity, specificity and area under curve of 98%, 96.66%,100% and 0.98 respectively. We have also developed a breast cancer risk index (BCRI) using significant KLPP features which can discriminate the two classes using a single integrated index. This can help the radiologists to discriminate the normal and malignant classes during screening to validate their findings.

Original languageEnglish
Pages (from-to)195-209
Number of pages15
JournalQuantitative InfraRed Thermography Journal
Volume13
Issue number2
DOIs
Publication statusPublished - 02-07-2016

Fingerprint

thermograms
histograms
breast
preserving
Screening
screening
projection
cancer
gradients
Tumors
tumors
Computer aided diagnosis
machine learning
Mammography
death
Learning algorithms
Learning systems
causes
sensitivity
curves

All Science Journal Classification (ASJC) codes

  • Instrumentation
  • Electrical and Electronic Engineering

Cite this

@article{28ece00b5e21459cb35e079bbd169f76,
title = "An integrated index for breast cancer identification using histogram of oriented gradient and kernel locality preserving projection features extracted from thermograms",
abstract = "Breast cancer is one of the prime causes of death in women worldwide. Thermography has shown a great potential in screening the breast cancer and overcomes the limitations of mammography. Moreover, interpretations of thermogram images are dependent on the specialists, which may lead to errors and uneven results. Preliminary screening method should detect the hazardous, destructive tumours effectively to improve the accuracy. The growth of malignant tumour can increase the internal temperature which can be captured by thermograms. Thus in this work, locally normalised histogram of oriented gradients (HOG) based preliminary screening computer aided diagnosis tool is proposed. HOG is able to record the minute internal variations in thermograms. In order to reduce the dimensions of extracted HOG descriptors kernel locality preserving projection (KLPP) is used. The resulting KLPP features are then ranked to form an efficient classification model. Various machine learning algorithms are used to validate the proposed method. Our method shows a promising performance with an average accuracy, sensitivity, specificity and area under curve of 98{\%}, 96.66{\%},100{\%} and 0.98 respectively. We have also developed a breast cancer risk index (BCRI) using significant KLPP features which can discriminate the two classes using a single integrated index. This can help the radiologists to discriminate the normal and malignant classes during screening to validate their findings.",
author = "U. Raghavendra and {Rajendra Acharya}, U. and Ng, {E. Y.K.} and Tan, {Jen Hong} and Anjan Gudigar",
year = "2016",
month = "7",
day = "2",
doi = "10.1080/17686733.2016.1176734",
language = "English",
volume = "13",
pages = "195--209",
journal = "Quantitative InfraRed Thermography Journal",
issn = "1768-6733",
publisher = "Taylor and Francis Ltd.",
number = "2",

}

TY - JOUR

T1 - An integrated index for breast cancer identification using histogram of oriented gradient and kernel locality preserving projection features extracted from thermograms

AU - Raghavendra, U.

AU - Rajendra Acharya, U.

AU - Ng, E. Y.K.

AU - Tan, Jen Hong

AU - Gudigar, Anjan

PY - 2016/7/2

Y1 - 2016/7/2

N2 - Breast cancer is one of the prime causes of death in women worldwide. Thermography has shown a great potential in screening the breast cancer and overcomes the limitations of mammography. Moreover, interpretations of thermogram images are dependent on the specialists, which may lead to errors and uneven results. Preliminary screening method should detect the hazardous, destructive tumours effectively to improve the accuracy. The growth of malignant tumour can increase the internal temperature which can be captured by thermograms. Thus in this work, locally normalised histogram of oriented gradients (HOG) based preliminary screening computer aided diagnosis tool is proposed. HOG is able to record the minute internal variations in thermograms. In order to reduce the dimensions of extracted HOG descriptors kernel locality preserving projection (KLPP) is used. The resulting KLPP features are then ranked to form an efficient classification model. Various machine learning algorithms are used to validate the proposed method. Our method shows a promising performance with an average accuracy, sensitivity, specificity and area under curve of 98%, 96.66%,100% and 0.98 respectively. We have also developed a breast cancer risk index (BCRI) using significant KLPP features which can discriminate the two classes using a single integrated index. This can help the radiologists to discriminate the normal and malignant classes during screening to validate their findings.

AB - Breast cancer is one of the prime causes of death in women worldwide. Thermography has shown a great potential in screening the breast cancer and overcomes the limitations of mammography. Moreover, interpretations of thermogram images are dependent on the specialists, which may lead to errors and uneven results. Preliminary screening method should detect the hazardous, destructive tumours effectively to improve the accuracy. The growth of malignant tumour can increase the internal temperature which can be captured by thermograms. Thus in this work, locally normalised histogram of oriented gradients (HOG) based preliminary screening computer aided diagnosis tool is proposed. HOG is able to record the minute internal variations in thermograms. In order to reduce the dimensions of extracted HOG descriptors kernel locality preserving projection (KLPP) is used. The resulting KLPP features are then ranked to form an efficient classification model. Various machine learning algorithms are used to validate the proposed method. Our method shows a promising performance with an average accuracy, sensitivity, specificity and area under curve of 98%, 96.66%,100% and 0.98 respectively. We have also developed a breast cancer risk index (BCRI) using significant KLPP features which can discriminate the two classes using a single integrated index. This can help the radiologists to discriminate the normal and malignant classes during screening to validate their findings.

UR - http://www.scopus.com/inward/record.url?scp=84965069050&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84965069050&partnerID=8YFLogxK

U2 - 10.1080/17686733.2016.1176734

DO - 10.1080/17686733.2016.1176734

M3 - Article

VL - 13

SP - 195

EP - 209

JO - Quantitative InfraRed Thermography Journal

JF - Quantitative InfraRed Thermography Journal

SN - 1768-6733

IS - 2

ER -