Impact of image compression on deep learning-based mammogram classification

Abstract Image compression is used in several clinical organizations to help address the overhead associated with medical imaging. These methods reduce file size by using a compact representation of the original image. This study aimed to analyze the impact of image compression on the performance of...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Yong-Yeon Jo, Young Sang Choi, Hyun Woo Park, Jae Hyeok Lee, Hyojung Jung, Hyo-Eun Kim, Kyounglan Ko, Chan Wha Lee, Hyo Soung Cha, Yul Hwangbo
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/62d819ec026a4b0db3bd394824a170e1
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:62d819ec026a4b0db3bd394824a170e1
record_format	dspace
spelling	oai:doaj.org-article:62d819ec026a4b0db3bd394824a170e12021-12-02T15:51:12ZImpact of image compression on deep learning-based mammogram classification10.1038/s41598-021-86726-w2045-2322https://doaj.org/article/62d819ec026a4b0db3bd394824a170e12021-04-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-86726-whttps://doaj.org/toc/2045-2322Abstract Image compression is used in several clinical organizations to help address the overhead associated with medical imaging. These methods reduce file size by using a compact representation of the original image. This study aimed to analyze the impact of image compression on the performance of deep learning-based models in classifying mammograms as “malignant”—cases that lead to a cancer diagnosis and treatment—or “normal” and “benign,” non-malignant cases that do not require immediate medical intervention. In this retrospective study, 9111 unique mammograms–5672 normal, 1686 benign, and 1754 malignant cases were collected from the National Cancer Center in the Republic of Korea. Image compression was applied to mammograms with compression ratios (CRs) ranging from 15 to 11 K. Convolutional neural networks (CNNs) with three convolutional layers and three fully-connected layers were trained using these images to classify a mammogram as malignant or not malignant across a range of CRs using five-fold cross-validation. Models trained on images with maximum CRs of 5 K had an average area under the receiver operating characteristic curve (AUROC) of 0.87 and area under the precision-recall curve (AUPRC) of 0.75 across the five folds and compression ratios. For images compressed with CRs of 10 K and 11 K, model performance decreased (average 0.79 in AUROC and 0.49 in AUPRC). Upon generating saliency maps that visualize the areas each model views as significant for prediction, models trained on less compressed (CR < = 5 K) images had maps encapsulating a radiologist’s label, while models trained on images with higher amounts of compression had maps that missed the ground truth completely. In addition, base ResNet18 models pre-trained on ImageNet and trained using compressed mammograms did not show performance improvements over our CNN model, with AUROC and AUPRC values ranging from 0.77 to 0.87 and 0.52 to 0.71 respectively when trained and tested on images with maximum CRs of 5 K. This paper finds that while training models on images with increased the robustness of the models when tested on compressed data, moderate image compression did not substantially impact the classification performance of DL-based models.Yong-Yeon JoYoung Sang ChoiHyun Woo ParkJae Hyeok LeeHyojung JungHyo-Eun KimKyounglan KoChan Wha LeeHyo Soung ChaYul HwangboNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-9 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Yong-Yeon Jo Young Sang Choi Hyun Woo Park Jae Hyeok Lee Hyojung Jung Hyo-Eun Kim Kyounglan Ko Chan Wha Lee Hyo Soung Cha Yul Hwangbo Impact of image compression on deep learning-based mammogram classification
description	Abstract Image compression is used in several clinical organizations to help address the overhead associated with medical imaging. These methods reduce file size by using a compact representation of the original image. This study aimed to analyze the impact of image compression on the performance of deep learning-based models in classifying mammograms as “malignant”—cases that lead to a cancer diagnosis and treatment—or “normal” and “benign,” non-malignant cases that do not require immediate medical intervention. In this retrospective study, 9111 unique mammograms–5672 normal, 1686 benign, and 1754 malignant cases were collected from the National Cancer Center in the Republic of Korea. Image compression was applied to mammograms with compression ratios (CRs) ranging from 15 to 11 K. Convolutional neural networks (CNNs) with three convolutional layers and three fully-connected layers were trained using these images to classify a mammogram as malignant or not malignant across a range of CRs using five-fold cross-validation. Models trained on images with maximum CRs of 5 K had an average area under the receiver operating characteristic curve (AUROC) of 0.87 and area under the precision-recall curve (AUPRC) of 0.75 across the five folds and compression ratios. For images compressed with CRs of 10 K and 11 K, model performance decreased (average 0.79 in AUROC and 0.49 in AUPRC). Upon generating saliency maps that visualize the areas each model views as significant for prediction, models trained on less compressed (CR < = 5 K) images had maps encapsulating a radiologist’s label, while models trained on images with higher amounts of compression had maps that missed the ground truth completely. In addition, base ResNet18 models pre-trained on ImageNet and trained using compressed mammograms did not show performance improvements over our CNN model, with AUROC and AUPRC values ranging from 0.77 to 0.87 and 0.52 to 0.71 respectively when trained and tested on images with maximum CRs of 5 K. This paper finds that while training models on images with increased the robustness of the models when tested on compressed data, moderate image compression did not substantially impact the classification performance of DL-based models.
format	article
author	Yong-Yeon Jo Young Sang Choi Hyun Woo Park Jae Hyeok Lee Hyojung Jung Hyo-Eun Kim Kyounglan Ko Chan Wha Lee Hyo Soung Cha Yul Hwangbo
author_facet	Yong-Yeon Jo Young Sang Choi Hyun Woo Park Jae Hyeok Lee Hyojung Jung Hyo-Eun Kim Kyounglan Ko Chan Wha Lee Hyo Soung Cha Yul Hwangbo
author_sort	Yong-Yeon Jo
title	Impact of image compression on deep learning-based mammogram classification
title_short	Impact of image compression on deep learning-based mammogram classification
title_full	Impact of image compression on deep learning-based mammogram classification
title_fullStr	Impact of image compression on deep learning-based mammogram classification
title_full_unstemmed	Impact of image compression on deep learning-based mammogram classification
title_sort	impact of image compression on deep learning-based mammogram classification
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/62d819ec026a4b0db3bd394824a170e1
work_keys_str_mv	AT yongyeonjo impactofimagecompressionondeeplearningbasedmammogramclassification AT youngsangchoi impactofimagecompressionondeeplearningbasedmammogramclassification AT hyunwoopark impactofimagecompressionondeeplearningbasedmammogramclassification AT jaehyeoklee impactofimagecompressionondeeplearningbasedmammogramclassification AT hyojungjung impactofimagecompressionondeeplearningbasedmammogramclassification AT hyoeunkim impactofimagecompressionondeeplearningbasedmammogramclassification AT kyounglanko impactofimagecompressionondeeplearningbasedmammogramclassification AT chanwhalee impactofimagecompressionondeeplearningbasedmammogramclassification AT hyosoungcha impactofimagecompressionondeeplearningbasedmammogramclassification AT yulhwangbo impactofimagecompressionondeeplearningbasedmammogramclassification
_version_	1718385635243327488

Impact of image compression on deep learning-based mammogram classification

Ejemplares similares