The effect of image resolution on convolutional neural networks in breast ultrasound

PURPOSE: The objective of this research was to investigate the efficacy of various parameter combinations of Convolutional Neural Networks (CNNs) models, namely MobileNet and DenseNet121, and different input image resolutions (REZs) ranging from 64×64 to 512×512 pixels, for diagnosing breast cancer....

Descripción completa

Detalles Bibliográficos
Autores principales: Tang, Shuzhen, Jing, Chen, Jiang, Yitao, Yang, Keen, Huang, Zhibin, Wu, Huaiyu, Cui, Chen, Shi, Siyuan, Ye, Xiuqin, Tian, Hongtian, Song, Di, Xu, Jinfeng, Dong, Fajin
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Elsevier 2023
Materias:
Research Article
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10469557/
https://www.ncbi.nlm.nih.gov/pubmed/37664701
http://dx.doi.org/10.1016/j.heliyon.2023.e19253
_version_ 1785099467820630016
author Tang, Shuzhen
Jing, Chen
Jiang, Yitao
Yang, Keen
Huang, Zhibin
Wu, Huaiyu
Cui, Chen
Shi, Siyuan
Ye, Xiuqin
Tian, Hongtian
Song, Di
Xu, Jinfeng
Dong, Fajin
author_facet Tang, Shuzhen
Jing, Chen
Jiang, Yitao
Yang, Keen
Huang, Zhibin
Wu, Huaiyu
Cui, Chen
Shi, Siyuan
Ye, Xiuqin
Tian, Hongtian
Song, Di
Xu, Jinfeng
Dong, Fajin
author_sort Tang, Shuzhen
collection PubMed
description PURPOSE: The objective of this research was to investigate the efficacy of various parameter combinations of Convolutional Neural Networks (CNNs) models, namely MobileNet and DenseNet121, and different input image resolutions (REZs) ranging from 64×64 to 512×512 pixels, for diagnosing breast cancer. MATERIALS AND METHODS: During the period of June 2015 to November 2020, two hospitals were involved in the collection of two-dimensional ultrasound breast images for this retrospective multicenter study. The diagnostic performance of the computer models MobileNet and DenseNet 121 was compared at different resolutions. RESULTS: The results showed that MobileNet had the best breast cancer diagnosis performance at 320×320pixel REZ and DenseNet121 had the best breast cancer diagnosis performance at 448×448pixel REZ. CONCLUSION: Our study reveals a significant correlation between image resolution and breast cancer diagnosis accuracy. Through the comparison of MobileNet and DenseNet121, it is highlighted that lightweight neural networks (LW-CNNs) can achieve model performance similar to or even slightly better than large neural networks models (HW-CNNs) in ultrasound images, and LW-CNNs' prediction time per image is lower.
format Online
Article
Text
id pubmed-10469557
institution National Center for Biotechnology Information
language English
publishDate 2023
publisher Elsevier
record_format MEDLINE/PubMed
spelling pubmed-104695572023-09-01 The effect of image resolution on convolutional neural networks in breast ultrasound Tang, Shuzhen Jing, Chen Jiang, Yitao Yang, Keen Huang, Zhibin Wu, Huaiyu Cui, Chen Shi, Siyuan Ye, Xiuqin Tian, Hongtian Song, Di Xu, Jinfeng Dong, Fajin Heliyon Research Article PURPOSE: The objective of this research was to investigate the efficacy of various parameter combinations of Convolutional Neural Networks (CNNs) models, namely MobileNet and DenseNet121, and different input image resolutions (REZs) ranging from 64×64 to 512×512 pixels, for diagnosing breast cancer. MATERIALS AND METHODS: During the period of June 2015 to November 2020, two hospitals were involved in the collection of two-dimensional ultrasound breast images for this retrospective multicenter study. The diagnostic performance of the computer models MobileNet and DenseNet 121 was compared at different resolutions. RESULTS: The results showed that MobileNet had the best breast cancer diagnosis performance at 320×320pixel REZ and DenseNet121 had the best breast cancer diagnosis performance at 448×448pixel REZ. CONCLUSION: Our study reveals a significant correlation between image resolution and breast cancer diagnosis accuracy. Through the comparison of MobileNet and DenseNet121, it is highlighted that lightweight neural networks (LW-CNNs) can achieve model performance similar to or even slightly better than large neural networks models (HW-CNNs) in ultrasound images, and LW-CNNs' prediction time per image is lower. Elsevier 2023-08-21 /pmc/articles/PMC10469557/ /pubmed/37664701 http://dx.doi.org/10.1016/j.heliyon.2023.e19253 Text en © 2023 The Authors https://creativecommons.org/licenses/by-nc-nd/4.0/This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
spellingShingle Research Article
Tang, Shuzhen
Jing, Chen
Jiang, Yitao
Yang, Keen
Huang, Zhibin
Wu, Huaiyu
Cui, Chen
Shi, Siyuan
Ye, Xiuqin
Tian, Hongtian
Song, Di
Xu, Jinfeng
Dong, Fajin
The effect of image resolution on convolutional neural networks in breast ultrasound
title The effect of image resolution on convolutional neural networks in breast ultrasound
title_full The effect of image resolution on convolutional neural networks in breast ultrasound
title_fullStr The effect of image resolution on convolutional neural networks in breast ultrasound
title_full_unstemmed The effect of image resolution on convolutional neural networks in breast ultrasound
title_short The effect of image resolution on convolutional neural networks in breast ultrasound
title_sort effect of image resolution on convolutional neural networks in breast ultrasound
topic Research Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10469557/
https://www.ncbi.nlm.nih.gov/pubmed/37664701
http://dx.doi.org/10.1016/j.heliyon.2023.e19253
work_keys_str_mv AT tangshuzhen theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT jingchen theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT jiangyitao theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT yangkeen theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT huangzhibin theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT wuhuaiyu theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT cuichen theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT shisiyuan theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT yexiuqin theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT tianhongtian theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT songdi theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT xujinfeng theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT dongfajin theeffectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT tangshuzhen effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT jingchen effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT jiangyitao effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT yangkeen effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT huangzhibin effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT wuhuaiyu effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT cuichen effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT shisiyuan effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT yexiuqin effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT tianhongtian effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT songdi effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT xujinfeng effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound
AT dongfajin effectofimageresolutiononconvolutionalneuralnetworksinbreastultrasound

Ejemplares similares