Pre-trained convolutional neural network with transfer learning by artificial illustrated images classify power Doppler ultrasound images of rheumatoid arthritis joints

Jun Fukae, Yoshiharu Amasaki, Yuichiro Fujieda, Yuki Sone, Ken Katagishi, Tatsunori Horie, Tamotsu Kamishima, Tatsuya Atsumi
{"title":"Pre-trained convolutional neural network with transfer learning by artificial illustrated images classify power Doppler ultrasound images of rheumatoid arthritis joints","authors":"Jun Fukae, Yoshiharu Amasaki, Yuichiro Fujieda, Yuki Sone, Ken Katagishi, Tatsunori Horie, Tamotsu Kamishima, Tatsuya Atsumi","doi":"10.1101/2024.08.30.24312848","DOIUrl":null,"url":null,"abstract":"This research aimed to study the classification performance of a pre-trained convolutional neural network (CNN) with transfer learning by artificial images of the joint ultrasonography in rheumatoid arthritis (RA). We focused on abnormal synovial vascularity and created 870 artificial ultrasound joint images based on the European League Against Rheumatism/Outcome Measure in Rheumatology scoring system. One CNN, the Visual Geometry Group (VGG)-16 was trained with transfer learning using the 870 artificial images for initial training and the original plus five additional images for second training. Actual joint ultrasound images obtained from patients with RA were used for testing our models. We obtained 156 actual ultrasound joint images from 74 patients with RA. Our initial model showed moderate classification performance, but grade 1 was especially low (area under curve (AUC) 0.59). In our second model, grade 1 showed improvement (AUC 0.73). We concluded that artificial images were useful for training VGG-16. Our novel approach of using artificial images as an alternative to actual images for training CNN has the potential to be applied in medical imaging fields that face difficulties in collecting real clinical images.","PeriodicalId":501212,"journal":{"name":"medRxiv - Rheumatology","volume":"185 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"medRxiv - Rheumatology","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1101/2024.08.30.24312848","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

This research aimed to study the classification performance of a pre-trained convolutional neural network (CNN) with transfer learning by artificial images of the joint ultrasonography in rheumatoid arthritis (RA). We focused on abnormal synovial vascularity and created 870 artificial ultrasound joint images based on the European League Against Rheumatism/Outcome Measure in Rheumatology scoring system. One CNN, the Visual Geometry Group (VGG)-16 was trained with transfer learning using the 870 artificial images for initial training and the original plus five additional images for second training. Actual joint ultrasound images obtained from patients with RA were used for testing our models. We obtained 156 actual ultrasound joint images from 74 patients with RA. Our initial model showed moderate classification performance, but grade 1 was especially low (area under curve (AUC) 0.59). In our second model, grade 1 showed improvement (AUC 0.73). We concluded that artificial images were useful for training VGG-16. Our novel approach of using artificial images as an alternative to actual images for training CNN has the potential to be applied in medical imaging fields that face difficulties in collecting real clinical images.
通过人工图解图像转移学习的预训练卷积神经网络对类风湿性关节炎关节的动力多普勒超声图像进行分类
本研究旨在通过类风湿性关节炎(RA)的人工关节超声图像,研究带有迁移学习的预训练卷积神经网络(CNN)的分类性能。我们重点研究了异常滑膜血管,并根据欧洲抗风湿病联盟/风湿病学成果测量评分系统创建了 870 幅人工超声关节图像。视觉几何组(VGG)-16 CNN 在初始训练中使用 870 幅人工图像进行迁移学习训练,在第二次训练中使用原始图像和另外五幅图像进行迁移学习训练。从 RA 患者处获得的实际关节超声图像用于测试我们的模型。我们从 74 名 RA 患者那里获得了 156 幅实际关节超声图像。我们的初始模型显示出中等的分类性能,但分级 1 尤其低(曲线下面积 (AUC) 0.59)。在我们的第二个模型中,等级 1 有所改善(AUC 0.73)。我们的结论是,人工图像对于训练 VGG-16 非常有用。我们使用人工图像替代真实图像来训练 CNN 的新方法有望应用于难以收集真实临床图像的医学影像领域。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信