AutoCovNet: Unsupervised feature learning using autoencoder and feature merging for detection of COVID-19 from chest X-ray images

With the onset of the COVID-19 pandemic, the automated diagnosis has become one of the most trending topics of research for faster mass screening. Deep learning-based approaches have been established as the most promising methods in this regard. However, the limitation of the labeled data is the mai...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Biocybernetics and biomedical engineering 2021-10, Vol.41 (4), p.1685-1701
Hauptverfasser: Rashid, Nayeeb, Hossain, Md Adnan Faisal, Ali, Mohammad, Islam Sukanya, Mumtahina, Mahmud, Tanvir, Fattah, Shaikh Anowarul
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 1701
container_issue 4
container_start_page 1685
container_title Biocybernetics and biomedical engineering
container_volume 41
creator Rashid, Nayeeb
Hossain, Md Adnan Faisal
Ali, Mohammad
Islam Sukanya, Mumtahina
Mahmud, Tanvir
Fattah, Shaikh Anowarul
description With the onset of the COVID-19 pandemic, the automated diagnosis has become one of the most trending topics of research for faster mass screening. Deep learning-based approaches have been established as the most promising methods in this regard. However, the limitation of the labeled data is the main bottleneck of the data-hungry deep learning methods. In this paper, a two-stage deep CNN based scheme is proposed to detect COVID-19 from chest X-ray images for achieving optimum performance with limited training images. In the first stage, an encoder-decoder based autoencoder network is proposed, trained on chest X-ray images in an unsupervised manner, and the network learns to reconstruct the X-ray images. An encoder-merging network is proposed for the second stage that consists of different layers of the encoder model followed by a merging network. Here the encoder model is initialized with the weights learned on the first stage and the outputs from different layers of the encoder model are used effectively by being connected to a proposed merging network. An intelligent feature merging scheme is introduced in the proposed merging network. Finally, the encoder-merging network is trained for feature extraction of the X-ray images in a supervised manner and resulting features are used in the classification layers of the proposed architecture. Considering the final classification task, an EfficientNet-B4 network is utilized in both stages. An end to end training is performed for datasets containing classes: COVID-19, Normal, Bacterial Pneumonia, Viral Pneumonia. The proposed method offers very satisfactory performances compared to the state of the art methods and achieves an accuracy of 90:13% on the 4-class, 96:45% on a 3-class, and 99:39% on 2-class classification.
doi_str_mv 10.1016/j.bbe.2021.09.004
format Article
fullrecord <record><control><sourceid>proquest_pubme</sourceid><recordid>TN_cdi_pubmedcentral_primary_oai_pubmedcentral_nih_gov_8526490</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><els_id>S020852162100108X</els_id><sourcerecordid>2585914373</sourcerecordid><originalsourceid>FETCH-LOGICAL-c451t-ad1beeba6d0c99a318dc19cde6c0b9ab37a2e9f2374cdd1f29fc749ca3eeaea03</originalsourceid><addsrcrecordid>eNp9kc1u1DAURi1ERUdtH4AN8pJNUv8kTgwSUjVQqFTRDUXdWY59PfUoiQc7GalL3hxHU0q7wQt7cY8_X9-D0FtKSkqoON-WXQclI4yWRJaEVK_QinFJi0o0d6_RijDSFjWj4hidpbQleQlaC87foGNeCUm4bFfo98U8hXXYf4fpA74d07yDuPcJLHagpzkC7kHH0Y8bPKdl15mH0QQLEevxHzZA3Cx1FyK2MIGZfBhxcHh98_Pqc0EldjEM2NxDmvBdEfUD9oPeQDpFR073Cc4ezxN0e_nlx_pbcX3z9Wp9cV2YqqZToS3tADotLDFSak5ba6g0FoQhndQdbzQD6RhvKmMtdUw601TSaA6gQRN-gj4dcndzN4A1ME5R92oXcxvxQQXt1cvK6O_VJuxVWzNRySXg_WNADL_m_A01-GSg7_UIYU6K1W0tacUbnlF6QE0MKUVwT89QohZ7aquyPbXYU0SqbC_fefe8v6cbf11l4OMBgDylvYeokvFZBVgf87iVDf4_8X8AtSqupw</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2585914373</pqid></control><display><type>article</type><title>AutoCovNet: Unsupervised feature learning using autoencoder and feature merging for detection of COVID-19 from chest X-ray images</title><source>Alma/SFX Local Collection</source><creator>Rashid, Nayeeb ; Hossain, Md Adnan Faisal ; Ali, Mohammad ; Islam Sukanya, Mumtahina ; Mahmud, Tanvir ; Fattah, Shaikh Anowarul</creator><creatorcontrib>Rashid, Nayeeb ; Hossain, Md Adnan Faisal ; Ali, Mohammad ; Islam Sukanya, Mumtahina ; Mahmud, Tanvir ; Fattah, Shaikh Anowarul</creatorcontrib><description>With the onset of the COVID-19 pandemic, the automated diagnosis has become one of the most trending topics of research for faster mass screening. Deep learning-based approaches have been established as the most promising methods in this regard. However, the limitation of the labeled data is the main bottleneck of the data-hungry deep learning methods. In this paper, a two-stage deep CNN based scheme is proposed to detect COVID-19 from chest X-ray images for achieving optimum performance with limited training images. In the first stage, an encoder-decoder based autoencoder network is proposed, trained on chest X-ray images in an unsupervised manner, and the network learns to reconstruct the X-ray images. An encoder-merging network is proposed for the second stage that consists of different layers of the encoder model followed by a merging network. Here the encoder model is initialized with the weights learned on the first stage and the outputs from different layers of the encoder model are used effectively by being connected to a proposed merging network. An intelligent feature merging scheme is introduced in the proposed merging network. Finally, the encoder-merging network is trained for feature extraction of the X-ray images in a supervised manner and resulting features are used in the classification layers of the proposed architecture. Considering the final classification task, an EfficientNet-B4 network is utilized in both stages. An end to end training is performed for datasets containing classes: COVID-19, Normal, Bacterial Pneumonia, Viral Pneumonia. The proposed method offers very satisfactory performances compared to the state of the art methods and achieves an accuracy of 90:13% on the 4-class, 96:45% on a 3-class, and 99:39% on 2-class classification.</description><identifier>ISSN: 0208-5216</identifier><identifier>EISSN: 2391-467X</identifier><identifier>DOI: 10.1016/j.bbe.2021.09.004</identifier><identifier>PMID: 34690398</identifier><language>eng</language><publisher>Poland: Elsevier B.V</publisher><subject>Autoencoder ; COVID-19 diagnosis ; Medical Image Analysis ; Neural Network ; Original ; X-ray</subject><ispartof>Biocybernetics and biomedical engineering, 2021-10, Vol.41 (4), p.1685-1701</ispartof><rights>2021 Nalecz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences</rights><rights>2021 Nalecz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences. Published by Elsevier B.V. All rights reserved.</rights><rights>2021 Nalecz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences. Published by Elsevier B.V. All rights reserved. 2021 Nalecz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c451t-ad1beeba6d0c99a318dc19cde6c0b9ab37a2e9f2374cdd1f29fc749ca3eeaea03</citedby><cites>FETCH-LOGICAL-c451t-ad1beeba6d0c99a318dc19cde6c0b9ab37a2e9f2374cdd1f29fc749ca3eeaea03</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>230,314,780,784,885,27924,27925</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/34690398$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Rashid, Nayeeb</creatorcontrib><creatorcontrib>Hossain, Md Adnan Faisal</creatorcontrib><creatorcontrib>Ali, Mohammad</creatorcontrib><creatorcontrib>Islam Sukanya, Mumtahina</creatorcontrib><creatorcontrib>Mahmud, Tanvir</creatorcontrib><creatorcontrib>Fattah, Shaikh Anowarul</creatorcontrib><title>AutoCovNet: Unsupervised feature learning using autoencoder and feature merging for detection of COVID-19 from chest X-ray images</title><title>Biocybernetics and biomedical engineering</title><addtitle>Biocybern Biomed Eng</addtitle><description>With the onset of the COVID-19 pandemic, the automated diagnosis has become one of the most trending topics of research for faster mass screening. Deep learning-based approaches have been established as the most promising methods in this regard. However, the limitation of the labeled data is the main bottleneck of the data-hungry deep learning methods. In this paper, a two-stage deep CNN based scheme is proposed to detect COVID-19 from chest X-ray images for achieving optimum performance with limited training images. In the first stage, an encoder-decoder based autoencoder network is proposed, trained on chest X-ray images in an unsupervised manner, and the network learns to reconstruct the X-ray images. An encoder-merging network is proposed for the second stage that consists of different layers of the encoder model followed by a merging network. Here the encoder model is initialized with the weights learned on the first stage and the outputs from different layers of the encoder model are used effectively by being connected to a proposed merging network. An intelligent feature merging scheme is introduced in the proposed merging network. Finally, the encoder-merging network is trained for feature extraction of the X-ray images in a supervised manner and resulting features are used in the classification layers of the proposed architecture. Considering the final classification task, an EfficientNet-B4 network is utilized in both stages. An end to end training is performed for datasets containing classes: COVID-19, Normal, Bacterial Pneumonia, Viral Pneumonia. The proposed method offers very satisfactory performances compared to the state of the art methods and achieves an accuracy of 90:13% on the 4-class, 96:45% on a 3-class, and 99:39% on 2-class classification.</description><subject>Autoencoder</subject><subject>COVID-19 diagnosis</subject><subject>Medical Image Analysis</subject><subject>Neural Network</subject><subject>Original</subject><subject>X-ray</subject><issn>0208-5216</issn><issn>2391-467X</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><recordid>eNp9kc1u1DAURi1ERUdtH4AN8pJNUv8kTgwSUjVQqFTRDUXdWY59PfUoiQc7GalL3hxHU0q7wQt7cY8_X9-D0FtKSkqoON-WXQclI4yWRJaEVK_QinFJi0o0d6_RijDSFjWj4hidpbQleQlaC87foGNeCUm4bFfo98U8hXXYf4fpA74d07yDuPcJLHagpzkC7kHH0Y8bPKdl15mH0QQLEevxHzZA3Cx1FyK2MIGZfBhxcHh98_Pqc0EldjEM2NxDmvBdEfUD9oPeQDpFR073Cc4ezxN0e_nlx_pbcX3z9Wp9cV2YqqZToS3tADotLDFSak5ba6g0FoQhndQdbzQD6RhvKmMtdUw601TSaA6gQRN-gj4dcndzN4A1ME5R92oXcxvxQQXt1cvK6O_VJuxVWzNRySXg_WNADL_m_A01-GSg7_UIYU6K1W0tacUbnlF6QE0MKUVwT89QohZ7aquyPbXYU0SqbC_fefe8v6cbf11l4OMBgDylvYeokvFZBVgf87iVDf4_8X8AtSqupw</recordid><startdate>20211001</startdate><enddate>20211001</enddate><creator>Rashid, Nayeeb</creator><creator>Hossain, Md Adnan Faisal</creator><creator>Ali, Mohammad</creator><creator>Islam Sukanya, Mumtahina</creator><creator>Mahmud, Tanvir</creator><creator>Fattah, Shaikh Anowarul</creator><general>Elsevier B.V</general><general>Nalecz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences. Published by Elsevier B.V</general><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7X8</scope><scope>5PM</scope></search><sort><creationdate>20211001</creationdate><title>AutoCovNet: Unsupervised feature learning using autoencoder and feature merging for detection of COVID-19 from chest X-ray images</title><author>Rashid, Nayeeb ; Hossain, Md Adnan Faisal ; Ali, Mohammad ; Islam Sukanya, Mumtahina ; Mahmud, Tanvir ; Fattah, Shaikh Anowarul</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c451t-ad1beeba6d0c99a318dc19cde6c0b9ab37a2e9f2374cdd1f29fc749ca3eeaea03</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Autoencoder</topic><topic>COVID-19 diagnosis</topic><topic>Medical Image Analysis</topic><topic>Neural Network</topic><topic>Original</topic><topic>X-ray</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Rashid, Nayeeb</creatorcontrib><creatorcontrib>Hossain, Md Adnan Faisal</creatorcontrib><creatorcontrib>Ali, Mohammad</creatorcontrib><creatorcontrib>Islam Sukanya, Mumtahina</creatorcontrib><creatorcontrib>Mahmud, Tanvir</creatorcontrib><creatorcontrib>Fattah, Shaikh Anowarul</creatorcontrib><collection>PubMed</collection><collection>CrossRef</collection><collection>MEDLINE - Academic</collection><collection>PubMed Central (Full Participant titles)</collection><jtitle>Biocybernetics and biomedical engineering</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Rashid, Nayeeb</au><au>Hossain, Md Adnan Faisal</au><au>Ali, Mohammad</au><au>Islam Sukanya, Mumtahina</au><au>Mahmud, Tanvir</au><au>Fattah, Shaikh Anowarul</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>AutoCovNet: Unsupervised feature learning using autoencoder and feature merging for detection of COVID-19 from chest X-ray images</atitle><jtitle>Biocybernetics and biomedical engineering</jtitle><addtitle>Biocybern Biomed Eng</addtitle><date>2021-10-01</date><risdate>2021</risdate><volume>41</volume><issue>4</issue><spage>1685</spage><epage>1701</epage><pages>1685-1701</pages><issn>0208-5216</issn><eissn>2391-467X</eissn><abstract>With the onset of the COVID-19 pandemic, the automated diagnosis has become one of the most trending topics of research for faster mass screening. Deep learning-based approaches have been established as the most promising methods in this regard. However, the limitation of the labeled data is the main bottleneck of the data-hungry deep learning methods. In this paper, a two-stage deep CNN based scheme is proposed to detect COVID-19 from chest X-ray images for achieving optimum performance with limited training images. In the first stage, an encoder-decoder based autoencoder network is proposed, trained on chest X-ray images in an unsupervised manner, and the network learns to reconstruct the X-ray images. An encoder-merging network is proposed for the second stage that consists of different layers of the encoder model followed by a merging network. Here the encoder model is initialized with the weights learned on the first stage and the outputs from different layers of the encoder model are used effectively by being connected to a proposed merging network. An intelligent feature merging scheme is introduced in the proposed merging network. Finally, the encoder-merging network is trained for feature extraction of the X-ray images in a supervised manner and resulting features are used in the classification layers of the proposed architecture. Considering the final classification task, an EfficientNet-B4 network is utilized in both stages. An end to end training is performed for datasets containing classes: COVID-19, Normal, Bacterial Pneumonia, Viral Pneumonia. The proposed method offers very satisfactory performances compared to the state of the art methods and achieves an accuracy of 90:13% on the 4-class, 96:45% on a 3-class, and 99:39% on 2-class classification.</abstract><cop>Poland</cop><pub>Elsevier B.V</pub><pmid>34690398</pmid><doi>10.1016/j.bbe.2021.09.004</doi><tpages>17</tpages><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier ISSN: 0208-5216
ispartof Biocybernetics and biomedical engineering, 2021-10, Vol.41 (4), p.1685-1701
issn 0208-5216
2391-467X
language eng
recordid cdi_pubmedcentral_primary_oai_pubmedcentral_nih_gov_8526490
source Alma/SFX Local Collection
subjects Autoencoder
COVID-19 diagnosis
Medical Image Analysis
Neural Network
Original
X-ray
title AutoCovNet: Unsupervised feature learning using autoencoder and feature merging for detection of COVID-19 from chest X-ray images
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-24T21%3A27%3A14IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_pubme&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=AutoCovNet:%20Unsupervised%20feature%20learning%20using%20autoencoder%20and%20feature%20merging%20for%20detection%20of%20COVID-19%20from%20chest%20X-ray%20images&rft.jtitle=Biocybernetics%20and%20biomedical%20engineering&rft.au=Rashid,%20Nayeeb&rft.date=2021-10-01&rft.volume=41&rft.issue=4&rft.spage=1685&rft.epage=1701&rft.pages=1685-1701&rft.issn=0208-5216&rft.eissn=2391-467X&rft_id=info:doi/10.1016/j.bbe.2021.09.004&rft_dat=%3Cproquest_pubme%3E2585914373%3C/proquest_pubme%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2585914373&rft_id=info:pmid/34690398&rft_els_id=S020852162100108X&rfr_iscdi=true