An extensive review of state-of-the-art transfer learning techniques used in medical imaging: Open issues and challenges

IF 2.1 Q3 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE Journal of Intelligent Systems Pub Date : 2022-01-01 DOI:10.1515/jisys-2022-0198
Abdulrahman Abbas Mukhlif, Belal Al-Khateeb, M. Mohammed
{"title":"An extensive review of state-of-the-art transfer learning techniques used in medical imaging: Open issues and challenges","authors":"Abdulrahman Abbas Mukhlif, Belal Al-Khateeb, M. Mohammed","doi":"10.1515/jisys-2022-0198","DOIUrl":null,"url":null,"abstract":"Abstract Deep learning techniques, which use a massive technology known as convolutional neural networks, have shown excellent results in a variety of areas, including image processing and interpretation. However, as the depth of these networks grows, so does the demand for a large amount of labeled data required to train these networks. In particular, the medical field suffers from a lack of images because the procedure for obtaining labeled medical images in the healthcare field is difficult, expensive, and requires specialized expertise to add labels to images. Moreover, the process may be prone to errors and time-consuming. Current research has revealed transfer learning as a viable solution to this problem. Transfer learning allows us to transfer knowledge gained from a previous process to improve and tackle a new problem. This study aims to conduct a comprehensive survey of recent studies that dealt with solving this problem and the most important metrics used to evaluate these methods. In addition, this study identifies problems in transfer learning techniques and highlights the problems of the medical dataset and potential problems that can be addressed in future research. According to our review, many researchers use pre-trained models on the Imagenet dataset (VGG16, ResNet, Inception v3) in many applications such as skin cancer, breast cancer, and diabetic retinopathy classification tasks. These techniques require further investigation of these models, due to training them on natural, non-medical images. In addition, many researchers use data augmentation techniques to expand their dataset and avoid overfitting. However, not enough studies have shown the effect of performance with or without data augmentation. Accuracy, recall, precision, F1 score, receiver operator characteristic curve, and area under the curve (AUC) were the most widely used measures in these studies. Furthermore, we identified problems in the datasets for melanoma and breast cancer and suggested corresponding solutions.","PeriodicalId":46139,"journal":{"name":"Journal of Intelligent Systems","volume":"33 1","pages":"1085 - 1111"},"PeriodicalIF":2.1000,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"9","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Intelligent Systems","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1515/jisys-2022-0198","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
引用次数: 9

Abstract

Abstract Deep learning techniques, which use a massive technology known as convolutional neural networks, have shown excellent results in a variety of areas, including image processing and interpretation. However, as the depth of these networks grows, so does the demand for a large amount of labeled data required to train these networks. In particular, the medical field suffers from a lack of images because the procedure for obtaining labeled medical images in the healthcare field is difficult, expensive, and requires specialized expertise to add labels to images. Moreover, the process may be prone to errors and time-consuming. Current research has revealed transfer learning as a viable solution to this problem. Transfer learning allows us to transfer knowledge gained from a previous process to improve and tackle a new problem. This study aims to conduct a comprehensive survey of recent studies that dealt with solving this problem and the most important metrics used to evaluate these methods. In addition, this study identifies problems in transfer learning techniques and highlights the problems of the medical dataset and potential problems that can be addressed in future research. According to our review, many researchers use pre-trained models on the Imagenet dataset (VGG16, ResNet, Inception v3) in many applications such as skin cancer, breast cancer, and diabetic retinopathy classification tasks. These techniques require further investigation of these models, due to training them on natural, non-medical images. In addition, many researchers use data augmentation techniques to expand their dataset and avoid overfitting. However, not enough studies have shown the effect of performance with or without data augmentation. Accuracy, recall, precision, F1 score, receiver operator characteristic curve, and area under the curve (AUC) were the most widely used measures in these studies. Furthermore, we identified problems in the datasets for melanoma and breast cancer and suggested corresponding solutions.
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
医学影像中使用的最先进的迁移学习技术的广泛回顾:开放的问题和挑战
深度学习技术使用了大量的卷积神经网络技术,在包括图像处理和解释在内的各个领域都取得了优异的成绩。然而,随着这些网络深度的增长,对训练这些网络所需的大量标记数据的需求也在增加。特别是,医疗领域缺乏图像,因为在医疗领域获得标记医学图像的过程困难,昂贵,并且需要专门的专业知识来为图像添加标签。此外,该过程可能容易出错且耗时。目前的研究表明,迁移学习是解决这一问题的可行方法。迁移学习允许我们将从以前的过程中获得的知识转移到改进和解决新问题。本研究旨在对最近解决这一问题的研究以及用于评估这些方法的最重要指标进行全面调查。此外,本研究确定了迁移学习技术中的问题,并强调了医学数据集的问题以及在未来研究中可以解决的潜在问题。根据我们的综述,许多研究人员使用Imagenet数据集(VGG16、ResNet、Inception v3)上的预训练模型进行皮肤癌、乳腺癌和糖尿病视网膜病变的分类任务。这些技术需要对这些模型进行进一步的研究,因为它们需要在自然的、非医学的图像上进行训练。此外,许多研究人员使用数据增强技术来扩展他们的数据集,避免过拟合。然而,并没有足够的研究表明数据增强或不增强对性能的影响。正确率、查全率、精密度、F1评分、接收者操作者特征曲线和曲线下面积(AUC)是这些研究中最广泛使用的测量指标。此外,我们发现了黑色素瘤和乳腺癌数据集中存在的问题,并提出了相应的解决方案。
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 去求助
来源期刊
Journal of Intelligent Systems
Journal of Intelligent Systems COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE-
CiteScore
5.90
自引率
3.30%
发文量
77
审稿时长
51 weeks
期刊介绍: The Journal of Intelligent Systems aims to provide research and review papers, as well as Brief Communications at an interdisciplinary level, with the field of intelligent systems providing the focal point. This field includes areas like artificial intelligence, models and computational theories of human cognition, perception and motivation; brain models, artificial neural nets and neural computing. It covers contributions from the social, human and computer sciences to the analysis and application of information technology.
期刊最新文献
A study on predicting crime rates through machine learning and data mining using text A multiorder feature tracking and explanation strategy for explainable deep learning Intelligent control system for industrial robots based on multi-source data fusion Reinforcement learning with Gaussian process regression using variational free energy A novel distance vector hop localization method for wireless sensor networks
×
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
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 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 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1