Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Machine learning algorithm for classification of breast ultrasound images

Karlsson, Jennie LU ; Ramkull, Jennifer ; Arvidsson, Ida LU ; Heyden, Anders LU orcid ; Åström, Kalle LU orcid ; Overgaard, Niels Christian LU and Lång, Kristina LU (2022) SPIE Medical Imaging 2022: 12033.
Abstract
Breast cancer is the most common type of cancer globally. Early detection is important for reducing the morbidity and mortality of breast cancer. The aim of this study was to evaluate the performance of different machine learning models to classify malignant or benign lesions on breast ultrasound images. Three different convolutional neural network approaches were implemented: (a) Simple convolutional neural network, (b) transfer learning using pre-trained InceptionV3, ResNet50V2, VGG19 and Xception and (c) deep feature networks based on combinations of the four transfer networks in (b). The data consisted of two breast ultrasound image data sets: (1) an open, single-vendor, data set collected by Cairo University at Baheya Hospital, Egypt,... (More)
Breast cancer is the most common type of cancer globally. Early detection is important for reducing the morbidity and mortality of breast cancer. The aim of this study was to evaluate the performance of different machine learning models to classify malignant or benign lesions on breast ultrasound images. Three different convolutional neural network approaches were implemented: (a) Simple convolutional neural network, (b) transfer learning using pre-trained InceptionV3, ResNet50V2, VGG19 and Xception and (c) deep feature networks based on combinations of the four transfer networks in (b). The data consisted of two breast ultrasound image data sets: (1) an open, single-vendor, data set collected by Cairo University at Baheya Hospital, Egypt, consisting of 437 benign lesions and 210 malignant lesions, where 10% was set to be a test set and the rest was used for training and validation (development) and (2) An in-house, multi-vendor data set collected at Unilabs Mammography Unit, Skåne University Hospital, Sweden, consisting of 13 benign lesions and 265 malignant lesions, was used as an external test set. Both test sets were used for evaluating the networks. The performance measures used were area under the receiver operating characteristic curve (AUC), sensitivity, specificity and weighted accuracy. Holdout, i.e. the splitting of the development data into training and validation data sets just once, was used to find a model with as good performance as possible. 10-fold cross-validation was also performed to provide uncertainty estimates. For the transfer networks which were obtained with holdout, Gradient-weighted Class Activation Mapping was used to generate heat maps indicating which part of the image contributed to the network’s decision. For 10-fold cross-validation it was possible to achieve a mean AUC of 92% and mean sensitivity of 95% for the transfer network based on Xception when testing on the first data set. When testing on the second data set it was possible to obtain a mean AUC of 75% and mean sensitivity of 86% for the combination of ResNet50V2 and Xception.
(Less)
Abstract (Swedish)
Breast cancer is the most common type of cancer globally. Early detection is important for reducing the morbidity and mortality of breast cancer. The aim of this study was to evaluate the performance of different machine learning models to classify malignant or benign lesions on breast ultrasound images. Three different convolutional neural network approaches were implemented: (a) Simple convolutional neural network, (b) transfer learning using pre-trained InceptionV3, ResNet50V2, VGG19 and Xception and (c) deep feature networks based on combinations of the four transfer networks in (b). The data consisted of two breast ultrasound image data sets: (1) an open, single-vendor, data set collected by Cairo University at Baheya Hospital, Egypt,... (More)
Breast cancer is the most common type of cancer globally. Early detection is important for reducing the morbidity and mortality of breast cancer. The aim of this study was to evaluate the performance of different machine learning models to classify malignant or benign lesions on breast ultrasound images. Three different convolutional neural network approaches were implemented: (a) Simple convolutional neural network, (b) transfer learning using pre-trained InceptionV3, ResNet50V2, VGG19 and Xception and (c) deep feature networks based on combinations of the four transfer networks in (b). The data consisted of two breast ultrasound image data sets: (1) an open, single-vendor, data set collected by Cairo University at Baheya Hospital, Egypt, consisting of 437 benign lesions and 210 malignant lesions, where 10% was set to be a test set and the rest was used for training and validation (development) and (2) An in-house, multi-vendor data set collected at Unilabs Mammography Unit, Skåne University Hospital, Sweden, consisting of 13 benign lesions and 265 malignant lesions, was used as an external test set. Both test sets were used for evaluating the networks. The performance measures used were area under the receiver operating characteristic curve (AUC), sensitivity, specificity and weighted accuracy. Holdout, i.e. the splitting of the development data into training and validation data sets just once, was used to find a model with as good performance as possible. 10-fold cross-validation was also performed to provide uncertainty estimates. For the transfer networks which were obtained with holdout, Gradient-weighted Class Activation Mapping was used to generate heat maps indicating which part of the image contributed to the network’s decision. For 10-fold cross-validation it was possible to achieve a mean AUC of 92% and mean sensitivity of 95% for the transfer network based on Xception when testing on the first data set. When testing on the second data set it was possible to obtain a mean AUC of 75% and mean sensitivity of 86% for the combination of ResNet50V2 and Xception. (Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; and
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
keywords
Breast Ultrasound Images, Breast Cancer, Convolutional Neural Networks, Transfer Learning
host publication
Medical Imaging 2022 : Computer-Aided Diagnosis - Computer-Aided Diagnosis
volume
12033
pages
11 pages
publisher
SPIE
conference name
SPIE Medical Imaging 2022:
conference location
San Diego, United States
conference dates
2022-02-20 - 2022-03-28
external identifiers
  • scopus:85132798158
ISBN
9781510649422
DOI
10.1117/12.2611824
project
23°N - Tropic of Cancer, accessible breast diagnostics for global health
language
English
LU publication?
yes
additional info
© (2022) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
id
8eb4e016-c21d-40ba-adde-7aab20ab6900
date added to LUP
2022-04-13 15:32:22
date last changed
2023-12-17 12:26:58
@inproceedings{8eb4e016-c21d-40ba-adde-7aab20ab6900,
  abstract     = {{Breast cancer is the most common type of cancer globally. Early detection is important for reducing the morbidity and mortality of breast cancer. The aim of this study was to evaluate the performance of different machine learning models to classify malignant or benign lesions on breast ultrasound images. Three different convolutional neural network approaches were implemented: (a) Simple convolutional neural network, (b) transfer learning using pre-trained InceptionV3, ResNet50V2, VGG19 and Xception and (c) deep feature networks based on combinations of the four transfer networks in (b). The data consisted of two breast ultrasound image data sets: (1) an open, single-vendor, data set collected by Cairo University at Baheya Hospital, Egypt, consisting of 437 benign lesions and 210 malignant lesions, where 10% was set to be a test set and the rest was used for training and validation (development) and (2) An in-house, multi-vendor data set collected at Unilabs Mammography Unit, Skåne University Hospital, Sweden, consisting of 13 benign lesions and 265 malignant lesions, was used as an external test set. Both test sets were used for evaluating the networks. The performance measures used were area under the receiver operating characteristic curve (AUC), sensitivity, specificity and weighted accuracy. Holdout, i.e. the splitting of the development data into training and validation data sets just once, was used to find a model with as good performance as possible. 10-fold cross-validation was also performed to provide uncertainty estimates. For the transfer networks which were obtained with holdout, Gradient-weighted Class Activation Mapping was used to generate heat maps indicating which part of the image contributed to the network’s decision. For 10-fold cross-validation it was possible to achieve a mean AUC of 92% and mean sensitivity of 95% for the transfer network based on Xception when testing on the first data set. When testing on the second data set it was possible to obtain a mean AUC of 75% and mean sensitivity of 86% for the combination of ResNet50V2 and Xception.<br/>}},
  author       = {{Karlsson, Jennie and Ramkull, Jennifer and Arvidsson, Ida and Heyden, Anders and Åström, Kalle and Overgaard, Niels Christian and Lång, Kristina}},
  booktitle    = {{Medical Imaging 2022 : Computer-Aided Diagnosis}},
  isbn         = {{9781510649422}},
  keywords     = {{Breast Ultrasound Images; Breast Cancer; Convolutional Neural Networks; Transfer Learning}},
  language     = {{eng}},
  month        = {{04}},
  publisher    = {{SPIE}},
  title        = {{Machine learning algorithm for classification of breast ultrasound images}},
  url          = {{http://dx.doi.org/10.1117/12.2611824}},
  doi          = {{10.1117/12.2611824}},
  volume       = {{12033}},
  year         = {{2022}},
}