Automatic microaneurysm detection in fundus image based on local cross-section transformation and multi-feature fusion

•A new microaneurysm detection method is presented using local cross-section transformation and multi-feature fusion.•A simple and effective method for microaneurysm candidate extraction is proposed based on local minimum region extraction and block filtering.•A series of descriptors is proposed bas...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Computer methods and programs in biomedicine 2020-11, Vol.196, p.105687-105687, Article 105687
Hauptverfasser: Du, Jingyu, Zou, Beiji, Chen, Changlong, Xu, Ziwen, Liu, Qing
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 105687
container_issue
container_start_page 105687
container_title Computer methods and programs in biomedicine
container_volume 196
creator Du, Jingyu
Zou, Beiji
Chen, Changlong
Xu, Ziwen
Liu, Qing
description •A new microaneurysm detection method is presented using local cross-section transformation and multi-feature fusion.•A simple and effective method for microaneurysm candidate extraction is proposed based on local minimum region extraction and block filtering.•A series of descriptors is proposed based on cross-section profile, local cross-section transformation, intensity and boundary, edge detector, and local saliency analysis. Background and objective: Retinal microaneurysm (MA) is one of the earliest clinical signs of diabetic retinopathy(DR). Its detection is essential for controlling DR and preventing vision loss. However, the spatial scale of MA is extremely small and the contrast to surrounding background is subtle, which make MA detection challenging. The purpose of this work is to automatically detect MAs from fundus images. Methods: Our MA detector involves two stages: MA candidate extraction and classification. In MA candidate extraction stage, local minimum region extraction and block filtering are used to exploit the regions where MA may exist. In this way, most of irrelavent background regions are discarded , which subsequently facilitates the training of MA classifier. In the second stage, multiple features are extracted to train the MA classifier. To distinguish MA from vascular regions, we propose a series of descriptors according to the cross-section profile of MA. Specially, as MAs are small and their contrast to surroundings is subtle, we propose local cross-section transformation (LCT) to amplify the difference between the MA and confusing structures. Finally, an under-sampling boosting-based classifier (RUSBoost) is trained to determine whether the candidate is an MA. Results: The proposed method is evaluated on three public available databases i.e. e-ophtha-MA, DiaretDB1 and ROC training set. It achieves high sensitivities for low false positive rates on the three databases. Using the FROC metric, the final scores are 0.516, 0.402 and 0.293 respectively, which are comparable to existing state-of-the-art methods. Conclusions: The proposed local cross-section transformation enhances the discrimination of descriptors by amplifying difference between MAs and confusing structures, which facilitates the classification and improves the detection performances. With the powerful descriptors, our method achieves state-of-the-art performances on three public datasets consistently.
doi_str_mv 10.1016/j.cmpb.2020.105687
format Article
fullrecord <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_proquest_miscellaneous_2437121072</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><els_id>S0169260720315200</els_id><sourcerecordid>2437121072</sourcerecordid><originalsourceid>FETCH-LOGICAL-c356t-dfc6fb8b8bba5af4041f23ce3e6da1674499073886d86847b59da7d10b22b3803</originalsourceid><addsrcrecordid>eNp9kEtrFTEYhoNY7Gn1D7iQLN3MMZfJZcBNKbUVCm7sOmSSL5LDXI65FPrvzXiOLksWIfme74X3QegjJXtKqPxy2Lv5OO4ZYduHkFq9QTuqFeuUkOIt2jVo6Jgk6hJd5XwghDAh5Dt0yZnmYhBqh55vallnW6LDc3RptQvU9JJn7KGAK3FdcFxwqIuvGcfZ_gI82gwet8G0OjvhtpRzl89wSXbJYU1bZHvaxeO5TiV2AWypCVpUboP36CLYKcOH832Nnr7d_bx96B5_3H-_vXnsHBeydD44GUbdzmiFDT3paWDcAQfpLZWq74eBKK619FrqXo1i8FZ5SkbGRq4Jv0afT7nHtP6ukIuZY3YwTa3nWrNhPVeUUaJYQ9kJ_VsoQTDH1AqnF0OJ2Xybg9l8m823OfluS5_O-XWcwf9f-Se4AV9PALSWzxGSyS7C4sDH1JQZv8bX8v8AjsuT1g</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2437121072</pqid></control><display><type>article</type><title>Automatic microaneurysm detection in fundus image based on local cross-section transformation and multi-feature fusion</title><source>MEDLINE</source><source>Access via ScienceDirect (Elsevier)</source><creator>Du, Jingyu ; Zou, Beiji ; Chen, Changlong ; Xu, Ziwen ; Liu, Qing</creator><creatorcontrib>Du, Jingyu ; Zou, Beiji ; Chen, Changlong ; Xu, Ziwen ; Liu, Qing</creatorcontrib><description>•A new microaneurysm detection method is presented using local cross-section transformation and multi-feature fusion.•A simple and effective method for microaneurysm candidate extraction is proposed based on local minimum region extraction and block filtering.•A series of descriptors is proposed based on cross-section profile, local cross-section transformation, intensity and boundary, edge detector, and local saliency analysis. Background and objective: Retinal microaneurysm (MA) is one of the earliest clinical signs of diabetic retinopathy(DR). Its detection is essential for controlling DR and preventing vision loss. However, the spatial scale of MA is extremely small and the contrast to surrounding background is subtle, which make MA detection challenging. The purpose of this work is to automatically detect MAs from fundus images. Methods: Our MA detector involves two stages: MA candidate extraction and classification. In MA candidate extraction stage, local minimum region extraction and block filtering are used to exploit the regions where MA may exist. In this way, most of irrelavent background regions are discarded , which subsequently facilitates the training of MA classifier. In the second stage, multiple features are extracted to train the MA classifier. To distinguish MA from vascular regions, we propose a series of descriptors according to the cross-section profile of MA. Specially, as MAs are small and their contrast to surroundings is subtle, we propose local cross-section transformation (LCT) to amplify the difference between the MA and confusing structures. Finally, an under-sampling boosting-based classifier (RUSBoost) is trained to determine whether the candidate is an MA. Results: The proposed method is evaluated on three public available databases i.e. e-ophtha-MA, DiaretDB1 and ROC training set. It achieves high sensitivities for low false positive rates on the three databases. Using the FROC metric, the final scores are 0.516, 0.402 and 0.293 respectively, which are comparable to existing state-of-the-art methods. Conclusions: The proposed local cross-section transformation enhances the discrimination of descriptors by amplifying difference between MAs and confusing structures, which facilitates the classification and improves the detection performances. With the powerful descriptors, our method achieves state-of-the-art performances on three public datasets consistently.</description><identifier>ISSN: 0169-2607</identifier><identifier>EISSN: 1872-7565</identifier><identifier>DOI: 10.1016/j.cmpb.2020.105687</identifier><identifier>PMID: 32835957</identifier><language>eng</language><publisher>Ireland: Elsevier B.V</publisher><subject>Algorithms ; Diabetic retinopathy ; Diabetic Retinopathy - diagnostic imaging ; Fundus images ; Fundus Oculi ; Humans ; Local cross-section transformation ; Medical image analysis ; Microaneurysm - diagnostic imaging ; Microaneurysm detection</subject><ispartof>Computer methods and programs in biomedicine, 2020-11, Vol.196, p.105687-105687, Article 105687</ispartof><rights>2020 Elsevier B.V.</rights><rights>Copyright © 2020 Elsevier B.V. All rights reserved.</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c356t-dfc6fb8b8bba5af4041f23ce3e6da1674499073886d86847b59da7d10b22b3803</citedby><cites>FETCH-LOGICAL-c356t-dfc6fb8b8bba5af4041f23ce3e6da1674499073886d86847b59da7d10b22b3803</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://dx.doi.org/10.1016/j.cmpb.2020.105687$$EHTML$$P50$$Gelsevier$$H</linktohtml><link.rule.ids>314,780,784,3550,27924,27925,45995</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/32835957$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Du, Jingyu</creatorcontrib><creatorcontrib>Zou, Beiji</creatorcontrib><creatorcontrib>Chen, Changlong</creatorcontrib><creatorcontrib>Xu, Ziwen</creatorcontrib><creatorcontrib>Liu, Qing</creatorcontrib><title>Automatic microaneurysm detection in fundus image based on local cross-section transformation and multi-feature fusion</title><title>Computer methods and programs in biomedicine</title><addtitle>Comput Methods Programs Biomed</addtitle><description>•A new microaneurysm detection method is presented using local cross-section transformation and multi-feature fusion.•A simple and effective method for microaneurysm candidate extraction is proposed based on local minimum region extraction and block filtering.•A series of descriptors is proposed based on cross-section profile, local cross-section transformation, intensity and boundary, edge detector, and local saliency analysis. Background and objective: Retinal microaneurysm (MA) is one of the earliest clinical signs of diabetic retinopathy(DR). Its detection is essential for controlling DR and preventing vision loss. However, the spatial scale of MA is extremely small and the contrast to surrounding background is subtle, which make MA detection challenging. The purpose of this work is to automatically detect MAs from fundus images. Methods: Our MA detector involves two stages: MA candidate extraction and classification. In MA candidate extraction stage, local minimum region extraction and block filtering are used to exploit the regions where MA may exist. In this way, most of irrelavent background regions are discarded , which subsequently facilitates the training of MA classifier. In the second stage, multiple features are extracted to train the MA classifier. To distinguish MA from vascular regions, we propose a series of descriptors according to the cross-section profile of MA. Specially, as MAs are small and their contrast to surroundings is subtle, we propose local cross-section transformation (LCT) to amplify the difference between the MA and confusing structures. Finally, an under-sampling boosting-based classifier (RUSBoost) is trained to determine whether the candidate is an MA. Results: The proposed method is evaluated on three public available databases i.e. e-ophtha-MA, DiaretDB1 and ROC training set. It achieves high sensitivities for low false positive rates on the three databases. Using the FROC metric, the final scores are 0.516, 0.402 and 0.293 respectively, which are comparable to existing state-of-the-art methods. Conclusions: The proposed local cross-section transformation enhances the discrimination of descriptors by amplifying difference between MAs and confusing structures, which facilitates the classification and improves the detection performances. With the powerful descriptors, our method achieves state-of-the-art performances on three public datasets consistently.</description><subject>Algorithms</subject><subject>Diabetic retinopathy</subject><subject>Diabetic Retinopathy - diagnostic imaging</subject><subject>Fundus images</subject><subject>Fundus Oculi</subject><subject>Humans</subject><subject>Local cross-section transformation</subject><subject>Medical image analysis</subject><subject>Microaneurysm - diagnostic imaging</subject><subject>Microaneurysm detection</subject><issn>0169-2607</issn><issn>1872-7565</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2020</creationdate><recordtype>article</recordtype><sourceid>EIF</sourceid><recordid>eNp9kEtrFTEYhoNY7Gn1D7iQLN3MMZfJZcBNKbUVCm7sOmSSL5LDXI65FPrvzXiOLksWIfme74X3QegjJXtKqPxy2Lv5OO4ZYduHkFq9QTuqFeuUkOIt2jVo6Jgk6hJd5XwghDAh5Dt0yZnmYhBqh55vallnW6LDc3RptQvU9JJn7KGAK3FdcFxwqIuvGcfZ_gI82gwet8G0OjvhtpRzl89wSXbJYU1bZHvaxeO5TiV2AWypCVpUboP36CLYKcOH832Nnr7d_bx96B5_3H-_vXnsHBeydD44GUbdzmiFDT3paWDcAQfpLZWq74eBKK619FrqXo1i8FZ5SkbGRq4Jv0afT7nHtP6ukIuZY3YwTa3nWrNhPVeUUaJYQ9kJ_VsoQTDH1AqnF0OJ2Xybg9l8m823OfluS5_O-XWcwf9f-Se4AV9PALSWzxGSyS7C4sDH1JQZv8bX8v8AjsuT1g</recordid><startdate>202011</startdate><enddate>202011</enddate><creator>Du, Jingyu</creator><creator>Zou, Beiji</creator><creator>Chen, Changlong</creator><creator>Xu, Ziwen</creator><creator>Liu, Qing</creator><general>Elsevier B.V</general><scope>CGR</scope><scope>CUY</scope><scope>CVF</scope><scope>ECM</scope><scope>EIF</scope><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7X8</scope></search><sort><creationdate>202011</creationdate><title>Automatic microaneurysm detection in fundus image based on local cross-section transformation and multi-feature fusion</title><author>Du, Jingyu ; Zou, Beiji ; Chen, Changlong ; Xu, Ziwen ; Liu, Qing</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c356t-dfc6fb8b8bba5af4041f23ce3e6da1674499073886d86847b59da7d10b22b3803</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Algorithms</topic><topic>Diabetic retinopathy</topic><topic>Diabetic Retinopathy - diagnostic imaging</topic><topic>Fundus images</topic><topic>Fundus Oculi</topic><topic>Humans</topic><topic>Local cross-section transformation</topic><topic>Medical image analysis</topic><topic>Microaneurysm - diagnostic imaging</topic><topic>Microaneurysm detection</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Du, Jingyu</creatorcontrib><creatorcontrib>Zou, Beiji</creatorcontrib><creatorcontrib>Chen, Changlong</creatorcontrib><creatorcontrib>Xu, Ziwen</creatorcontrib><creatorcontrib>Liu, Qing</creatorcontrib><collection>Medline</collection><collection>MEDLINE</collection><collection>MEDLINE (Ovid)</collection><collection>MEDLINE</collection><collection>MEDLINE</collection><collection>PubMed</collection><collection>CrossRef</collection><collection>MEDLINE - Academic</collection><jtitle>Computer methods and programs in biomedicine</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Du, Jingyu</au><au>Zou, Beiji</au><au>Chen, Changlong</au><au>Xu, Ziwen</au><au>Liu, Qing</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Automatic microaneurysm detection in fundus image based on local cross-section transformation and multi-feature fusion</atitle><jtitle>Computer methods and programs in biomedicine</jtitle><addtitle>Comput Methods Programs Biomed</addtitle><date>2020-11</date><risdate>2020</risdate><volume>196</volume><spage>105687</spage><epage>105687</epage><pages>105687-105687</pages><artnum>105687</artnum><issn>0169-2607</issn><eissn>1872-7565</eissn><abstract>•A new microaneurysm detection method is presented using local cross-section transformation and multi-feature fusion.•A simple and effective method for microaneurysm candidate extraction is proposed based on local minimum region extraction and block filtering.•A series of descriptors is proposed based on cross-section profile, local cross-section transformation, intensity and boundary, edge detector, and local saliency analysis. Background and objective: Retinal microaneurysm (MA) is one of the earliest clinical signs of diabetic retinopathy(DR). Its detection is essential for controlling DR and preventing vision loss. However, the spatial scale of MA is extremely small and the contrast to surrounding background is subtle, which make MA detection challenging. The purpose of this work is to automatically detect MAs from fundus images. Methods: Our MA detector involves two stages: MA candidate extraction and classification. In MA candidate extraction stage, local minimum region extraction and block filtering are used to exploit the regions where MA may exist. In this way, most of irrelavent background regions are discarded , which subsequently facilitates the training of MA classifier. In the second stage, multiple features are extracted to train the MA classifier. To distinguish MA from vascular regions, we propose a series of descriptors according to the cross-section profile of MA. Specially, as MAs are small and their contrast to surroundings is subtle, we propose local cross-section transformation (LCT) to amplify the difference between the MA and confusing structures. Finally, an under-sampling boosting-based classifier (RUSBoost) is trained to determine whether the candidate is an MA. Results: The proposed method is evaluated on three public available databases i.e. e-ophtha-MA, DiaretDB1 and ROC training set. It achieves high sensitivities for low false positive rates on the three databases. Using the FROC metric, the final scores are 0.516, 0.402 and 0.293 respectively, which are comparable to existing state-of-the-art methods. Conclusions: The proposed local cross-section transformation enhances the discrimination of descriptors by amplifying difference between MAs and confusing structures, which facilitates the classification and improves the detection performances. With the powerful descriptors, our method achieves state-of-the-art performances on three public datasets consistently.</abstract><cop>Ireland</cop><pub>Elsevier B.V</pub><pmid>32835957</pmid><doi>10.1016/j.cmpb.2020.105687</doi><tpages>1</tpages></addata></record>
fulltext fulltext
identifier ISSN: 0169-2607
ispartof Computer methods and programs in biomedicine, 2020-11, Vol.196, p.105687-105687, Article 105687
issn 0169-2607
1872-7565
language eng
recordid cdi_proquest_miscellaneous_2437121072
source MEDLINE; Access via ScienceDirect (Elsevier)
subjects Algorithms
Diabetic retinopathy
Diabetic Retinopathy - diagnostic imaging
Fundus images
Fundus Oculi
Humans
Local cross-section transformation
Medical image analysis
Microaneurysm - diagnostic imaging
Microaneurysm detection
title Automatic microaneurysm detection in fundus image based on local cross-section transformation and multi-feature fusion
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-24T20%3A37%3A22IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Automatic%20microaneurysm%20detection%20in%20fundus%20image%20based%20on%20local%20cross-section%20transformation%20and%20multi-feature%20fusion&rft.jtitle=Computer%20methods%20and%20programs%20in%20biomedicine&rft.au=Du,%20Jingyu&rft.date=2020-11&rft.volume=196&rft.spage=105687&rft.epage=105687&rft.pages=105687-105687&rft.artnum=105687&rft.issn=0169-2607&rft.eissn=1872-7565&rft_id=info:doi/10.1016/j.cmpb.2020.105687&rft_dat=%3Cproquest_cross%3E2437121072%3C/proquest_cross%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2437121072&rft_id=info:pmid/32835957&rft_els_id=S0169260720315200&rfr_iscdi=true