Computer-aided classification of breast masses using contrast-enhanced digital mammograms

Gopichandh Danala, Faranak Aghaei, Morteza Heidari, Teresa Wu, Bhavika Patel, Bin Zheng

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

3 Citations (Scopus)

Abstract

By taking advantages of both mammography and breast MRI, contrast-enhanced digital mammography (CEDM) has emerged as a new promising imaging modality to improve efficacy of breast cancer screening and diagnosis. The primary objective of study is to develop and evaluate a new computer-aided detection and diagnosis (CAD) scheme of CEDM images to classify between malignant and benign breast masses. A CEDM dataset consisting of 111 patients (33 benign and 78 malignant) was retrospectively assembled. Each case includes two types of images namely, low-energy (LE) and dual-energy subtracted (DES) images. First, CAD scheme applied a hybrid segmentation method to automatically segment masses depicting on LE and DES images separately. Optimal segmentation results from DES images were also mapped to LE images and vice versa. Next, a set of 109 quantitative image features related to mass shape and density heterogeneity was initially computed. Last, four multilayer perceptron-based machine learning classifiers integrated with correlationbased feature subset evaluator and leave-one-case-out cross-validation method was built to classify mass regions depicting on LE and DES images, respectively. Initially, when CAD scheme was applied to original segmentation of DES and LE images, the areas under ROC curves were 0.7585±0.0526 and 0.7534±0.0470, respectively. After optimal segmentation mapping from DES to LE images, AUC value of CAD scheme significantly increased to 0.8477±0.0376 (p<0.01). Since DES images eliminate overlapping effect of dense breast tissue on lesions, segmentation accuracy was significantly improved as compared to regular mammograms, the study demonstrated that computer-aided classification of breast masses using CEDM images yielded higher performance.

Original languageEnglish (US)
Title of host publicationMedical Imaging 2018
Subtitle of host publicationComputer-Aided Diagnosis
EditorsKensaku Mori, Nicholas Petrick
PublisherSPIE
Volume10575
ISBN (Electronic)9781510616394
DOIs
StatePublished - Jan 1 2018
EventMedical Imaging 2018: Computer-Aided Diagnosis - Houston, United States
Duration: Feb 12 2018Feb 15 2018

Other

OtherMedical Imaging 2018: Computer-Aided Diagnosis
CountryUnited States
CityHouston
Period2/12/182/15/18

Fingerprint

Mammography
breast
Breast
Area Under Curve
energy
Neural Networks (Computer)
Multilayer neural networks
Early Detection of Cancer
ROC Curve
Magnetic resonance imaging
Learning systems
Screening
Classifiers
Tissue
Breast Neoplasms
Imaging techniques
machine learning
self organizing systems
classifiers
lesions

Keywords

  • breast mass classification
  • computer-aided diagnosis (CAD)
  • Contrast-enhanced digital mammography (CEDM)
  • correlationbased feature subset evaluator
  • Dual-energy subtracted (DES) image
  • Low energy (LE) image
  • Mammography
  • multilayer perceptron

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Biomaterials
  • Atomic and Molecular Physics, and Optics
  • Radiology Nuclear Medicine and imaging

Cite this

Danala, G., Aghaei, F., Heidari, M., Wu, T., Patel, B., & Zheng, B. (2018). Computer-aided classification of breast masses using contrast-enhanced digital mammograms. In K. Mori, & N. Petrick (Eds.), Medical Imaging 2018: Computer-Aided Diagnosis (Vol. 10575). [105752K] SPIE. https://doi.org/10.1117/12.2293136

Computer-aided classification of breast masses using contrast-enhanced digital mammograms. / Danala, Gopichandh; Aghaei, Faranak; Heidari, Morteza; Wu, Teresa; Patel, Bhavika; Zheng, Bin.

Medical Imaging 2018: Computer-Aided Diagnosis. ed. / Kensaku Mori; Nicholas Petrick. Vol. 10575 SPIE, 2018. 105752K.

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

Danala, G, Aghaei, F, Heidari, M, Wu, T, Patel, B & Zheng, B 2018, Computer-aided classification of breast masses using contrast-enhanced digital mammograms. in K Mori & N Petrick (eds), Medical Imaging 2018: Computer-Aided Diagnosis. vol. 10575, 105752K, SPIE, Medical Imaging 2018: Computer-Aided Diagnosis, Houston, United States, 2/12/18. https://doi.org/10.1117/12.2293136
Danala G, Aghaei F, Heidari M, Wu T, Patel B, Zheng B. Computer-aided classification of breast masses using contrast-enhanced digital mammograms. In Mori K, Petrick N, editors, Medical Imaging 2018: Computer-Aided Diagnosis. Vol. 10575. SPIE. 2018. 105752K https://doi.org/10.1117/12.2293136
Danala, Gopichandh ; Aghaei, Faranak ; Heidari, Morteza ; Wu, Teresa ; Patel, Bhavika ; Zheng, Bin. / Computer-aided classification of breast masses using contrast-enhanced digital mammograms. Medical Imaging 2018: Computer-Aided Diagnosis. editor / Kensaku Mori ; Nicholas Petrick. Vol. 10575 SPIE, 2018.
@inproceedings{565db77319394c6e96a8bdba315da128,
title = "Computer-aided classification of breast masses using contrast-enhanced digital mammograms",
abstract = "By taking advantages of both mammography and breast MRI, contrast-enhanced digital mammography (CEDM) has emerged as a new promising imaging modality to improve efficacy of breast cancer screening and diagnosis. The primary objective of study is to develop and evaluate a new computer-aided detection and diagnosis (CAD) scheme of CEDM images to classify between malignant and benign breast masses. A CEDM dataset consisting of 111 patients (33 benign and 78 malignant) was retrospectively assembled. Each case includes two types of images namely, low-energy (LE) and dual-energy subtracted (DES) images. First, CAD scheme applied a hybrid segmentation method to automatically segment masses depicting on LE and DES images separately. Optimal segmentation results from DES images were also mapped to LE images and vice versa. Next, a set of 109 quantitative image features related to mass shape and density heterogeneity was initially computed. Last, four multilayer perceptron-based machine learning classifiers integrated with correlationbased feature subset evaluator and leave-one-case-out cross-validation method was built to classify mass regions depicting on LE and DES images, respectively. Initially, when CAD scheme was applied to original segmentation of DES and LE images, the areas under ROC curves were 0.7585±0.0526 and 0.7534±0.0470, respectively. After optimal segmentation mapping from DES to LE images, AUC value of CAD scheme significantly increased to 0.8477±0.0376 (p<0.01). Since DES images eliminate overlapping effect of dense breast tissue on lesions, segmentation accuracy was significantly improved as compared to regular mammograms, the study demonstrated that computer-aided classification of breast masses using CEDM images yielded higher performance.",
keywords = "breast mass classification, computer-aided diagnosis (CAD), Contrast-enhanced digital mammography (CEDM), correlationbased feature subset evaluator, Dual-energy subtracted (DES) image, Low energy (LE) image, Mammography, multilayer perceptron",
author = "Gopichandh Danala and Faranak Aghaei and Morteza Heidari and Teresa Wu and Bhavika Patel and Bin Zheng",
year = "2018",
month = "1",
day = "1",
doi = "10.1117/12.2293136",
language = "English (US)",
volume = "10575",
editor = "Kensaku Mori and Nicholas Petrick",
booktitle = "Medical Imaging 2018",
publisher = "SPIE",

}

TY - GEN

T1 - Computer-aided classification of breast masses using contrast-enhanced digital mammograms

AU - Danala, Gopichandh

AU - Aghaei, Faranak

AU - Heidari, Morteza

AU - Wu, Teresa

AU - Patel, Bhavika

AU - Zheng, Bin

PY - 2018/1/1

Y1 - 2018/1/1

N2 - By taking advantages of both mammography and breast MRI, contrast-enhanced digital mammography (CEDM) has emerged as a new promising imaging modality to improve efficacy of breast cancer screening and diagnosis. The primary objective of study is to develop and evaluate a new computer-aided detection and diagnosis (CAD) scheme of CEDM images to classify between malignant and benign breast masses. A CEDM dataset consisting of 111 patients (33 benign and 78 malignant) was retrospectively assembled. Each case includes two types of images namely, low-energy (LE) and dual-energy subtracted (DES) images. First, CAD scheme applied a hybrid segmentation method to automatically segment masses depicting on LE and DES images separately. Optimal segmentation results from DES images were also mapped to LE images and vice versa. Next, a set of 109 quantitative image features related to mass shape and density heterogeneity was initially computed. Last, four multilayer perceptron-based machine learning classifiers integrated with correlationbased feature subset evaluator and leave-one-case-out cross-validation method was built to classify mass regions depicting on LE and DES images, respectively. Initially, when CAD scheme was applied to original segmentation of DES and LE images, the areas under ROC curves were 0.7585±0.0526 and 0.7534±0.0470, respectively. After optimal segmentation mapping from DES to LE images, AUC value of CAD scheme significantly increased to 0.8477±0.0376 (p<0.01). Since DES images eliminate overlapping effect of dense breast tissue on lesions, segmentation accuracy was significantly improved as compared to regular mammograms, the study demonstrated that computer-aided classification of breast masses using CEDM images yielded higher performance.

AB - By taking advantages of both mammography and breast MRI, contrast-enhanced digital mammography (CEDM) has emerged as a new promising imaging modality to improve efficacy of breast cancer screening and diagnosis. The primary objective of study is to develop and evaluate a new computer-aided detection and diagnosis (CAD) scheme of CEDM images to classify between malignant and benign breast masses. A CEDM dataset consisting of 111 patients (33 benign and 78 malignant) was retrospectively assembled. Each case includes two types of images namely, low-energy (LE) and dual-energy subtracted (DES) images. First, CAD scheme applied a hybrid segmentation method to automatically segment masses depicting on LE and DES images separately. Optimal segmentation results from DES images were also mapped to LE images and vice versa. Next, a set of 109 quantitative image features related to mass shape and density heterogeneity was initially computed. Last, four multilayer perceptron-based machine learning classifiers integrated with correlationbased feature subset evaluator and leave-one-case-out cross-validation method was built to classify mass regions depicting on LE and DES images, respectively. Initially, when CAD scheme was applied to original segmentation of DES and LE images, the areas under ROC curves were 0.7585±0.0526 and 0.7534±0.0470, respectively. After optimal segmentation mapping from DES to LE images, AUC value of CAD scheme significantly increased to 0.8477±0.0376 (p<0.01). Since DES images eliminate overlapping effect of dense breast tissue on lesions, segmentation accuracy was significantly improved as compared to regular mammograms, the study demonstrated that computer-aided classification of breast masses using CEDM images yielded higher performance.

KW - breast mass classification

KW - computer-aided diagnosis (CAD)

KW - Contrast-enhanced digital mammography (CEDM)

KW - correlationbased feature subset evaluator

KW - Dual-energy subtracted (DES) image

KW - Low energy (LE) image

KW - Mammography

KW - multilayer perceptron

UR - http://www.scopus.com/inward/record.url?scp=85046272409&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046272409&partnerID=8YFLogxK

U2 - 10.1117/12.2293136

DO - 10.1117/12.2293136

M3 - Conference contribution

VL - 10575

BT - Medical Imaging 2018

A2 - Mori, Kensaku

A2 - Petrick, Nicholas

PB - SPIE

ER -