Mu-net: Multi-scale U-net for two-photon microscopy image denoising and restoration
Advances in two two-photon microscopy (2PM) have made three-dimensional (3D) neural imaging of deep cortical regions possible. However, 2PM often suffers from poor image quality because of various noise factors, including blur, white noise, and photo bleaching. In addition, the effectiveness of the...
Gespeichert in:
Veröffentlicht in: | Neural networks 2020-05, Vol.125, p.92-103 |
---|---|
Hauptverfasser: | , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | 103 |
---|---|
container_issue | |
container_start_page | 92 |
container_title | Neural networks |
container_volume | 125 |
creator | Lee, Sehyung Negishi, Makiko Urakubo, Hidetoshi Kasai, Haruo Ishii, Shin |
description | Advances in two two-photon microscopy (2PM) have made three-dimensional (3D) neural imaging of deep cortical regions possible. However, 2PM often suffers from poor image quality because of various noise factors, including blur, white noise, and photo bleaching. In addition, the effectiveness of the existing image processing methods is limited because of the special features of 2PM images such as deeper tissue penetration but higher image noises owing to rapid laser scanning. To address the denoising problems in 2PM 3D images, we present a new algorithm based on deep convolutional neural networks (CNNs). The proposed model consists of multiple U-nets in which an individual U-net removes noises at different scales and then yields a performance improvement based on a coarse-to-fine strategy. Moreover, the constituent CNNs employ fully 3D convolution operations. Such an architecture enables the proposed model to facilitate end-to-end learning without any pre/post processing. Based on the experiments on 2PM image denoising, we observed that our new algorithm demonstrates substantial performance improvements over other baseline methods. |
doi_str_mv | 10.1016/j.neunet.2020.01.026 |
format | Article |
fullrecord | <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_proquest_miscellaneous_2363071399</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><els_id>S0893608020300368</els_id><sourcerecordid>2363071399</sourcerecordid><originalsourceid>FETCH-LOGICAL-c474t-337684f8b3e1de36cff771ae0adb43b0a6ece6e6358f7948a1c08d2b97238ce23</originalsourceid><addsrcrecordid>eNp9kE9r3DAQxUVoyG7SfINSdOzFzkhyJbmHQlmaP7AhhzRnIcvjVItX2kp2Sr59tOy2x5wGhvfmzfsR8olBzYDJq00dcA441Rw41MBq4PKELJlWbcWV5h_IEnQrKgkaFuQ85w0ASN2IM7IQHJRuZbMkj_dzVY58o_fzOPkqOzsifdqv6BATnf7Gavc7TjHQrXcpZhd3r9Rv7TPSHkP02YdnakNPE-YpJjv5GD6S08GOGS-P84I8Xf_8tbqt1g83d6sf68o1qpkqIVR5Z9CdQNajkG4YlGIWwfZdIzqwEh1KlOKrHlTbaMsc6J53reJCO-Tignw53N2l-Gcu-Wbrs8NxtAHjnA0XUoBiom2LtDlI9x1ywsHsUmmRXg0Ds8dpNuaA0-xxGmCm4Cy2z8eEudti_9_0j18RfD8IsPR88ZhMdh6Dw94ndJPpo38_4Q1p-oja</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2363071399</pqid></control><display><type>article</type><title>Mu-net: Multi-scale U-net for two-photon microscopy image denoising and restoration</title><source>Access via ScienceDirect (Elsevier)</source><creator>Lee, Sehyung ; Negishi, Makiko ; Urakubo, Hidetoshi ; Kasai, Haruo ; Ishii, Shin</creator><creatorcontrib>Lee, Sehyung ; Negishi, Makiko ; Urakubo, Hidetoshi ; Kasai, Haruo ; Ishii, Shin</creatorcontrib><description>Advances in two two-photon microscopy (2PM) have made three-dimensional (3D) neural imaging of deep cortical regions possible. However, 2PM often suffers from poor image quality because of various noise factors, including blur, white noise, and photo bleaching. In addition, the effectiveness of the existing image processing methods is limited because of the special features of 2PM images such as deeper tissue penetration but higher image noises owing to rapid laser scanning. To address the denoising problems in 2PM 3D images, we present a new algorithm based on deep convolutional neural networks (CNNs). The proposed model consists of multiple U-nets in which an individual U-net removes noises at different scales and then yields a performance improvement based on a coarse-to-fine strategy. Moreover, the constituent CNNs employ fully 3D convolution operations. Such an architecture enables the proposed model to facilitate end-to-end learning without any pre/post processing. Based on the experiments on 2PM image denoising, we observed that our new algorithm demonstrates substantial performance improvements over other baseline methods.</description><identifier>ISSN: 0893-6080</identifier><identifier>EISSN: 1879-2782</identifier><identifier>DOI: 10.1016/j.neunet.2020.01.026</identifier><identifier>PMID: 32078964</identifier><language>eng</language><publisher>United States: Elsevier Ltd</publisher><subject>Deep learning ; GAN ; Image denoising ; Two-photon microscopy image ; U-net</subject><ispartof>Neural networks, 2020-05, Vol.125, p.92-103</ispartof><rights>2020 Elsevier Ltd</rights><rights>Copyright © 2020 Elsevier Ltd. All rights reserved.</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c474t-337684f8b3e1de36cff771ae0adb43b0a6ece6e6358f7948a1c08d2b97238ce23</citedby><cites>FETCH-LOGICAL-c474t-337684f8b3e1de36cff771ae0adb43b0a6ece6e6358f7948a1c08d2b97238ce23</cites><orcidid>0000-0003-2327-9027 ; 0000-0001-7137-4622 ; 0000-0002-1816-0040</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://dx.doi.org/10.1016/j.neunet.2020.01.026$$EHTML$$P50$$Gelsevier$$H</linktohtml><link.rule.ids>314,780,784,3550,27924,27925,45995</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/32078964$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Lee, Sehyung</creatorcontrib><creatorcontrib>Negishi, Makiko</creatorcontrib><creatorcontrib>Urakubo, Hidetoshi</creatorcontrib><creatorcontrib>Kasai, Haruo</creatorcontrib><creatorcontrib>Ishii, Shin</creatorcontrib><title>Mu-net: Multi-scale U-net for two-photon microscopy image denoising and restoration</title><title>Neural networks</title><addtitle>Neural Netw</addtitle><description>Advances in two two-photon microscopy (2PM) have made three-dimensional (3D) neural imaging of deep cortical regions possible. However, 2PM often suffers from poor image quality because of various noise factors, including blur, white noise, and photo bleaching. In addition, the effectiveness of the existing image processing methods is limited because of the special features of 2PM images such as deeper tissue penetration but higher image noises owing to rapid laser scanning. To address the denoising problems in 2PM 3D images, we present a new algorithm based on deep convolutional neural networks (CNNs). The proposed model consists of multiple U-nets in which an individual U-net removes noises at different scales and then yields a performance improvement based on a coarse-to-fine strategy. Moreover, the constituent CNNs employ fully 3D convolution operations. Such an architecture enables the proposed model to facilitate end-to-end learning without any pre/post processing. Based on the experiments on 2PM image denoising, we observed that our new algorithm demonstrates substantial performance improvements over other baseline methods.</description><subject>Deep learning</subject><subject>GAN</subject><subject>Image denoising</subject><subject>Two-photon microscopy image</subject><subject>U-net</subject><issn>0893-6080</issn><issn>1879-2782</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2020</creationdate><recordtype>article</recordtype><recordid>eNp9kE9r3DAQxUVoyG7SfINSdOzFzkhyJbmHQlmaP7AhhzRnIcvjVItX2kp2Sr59tOy2x5wGhvfmzfsR8olBzYDJq00dcA441Rw41MBq4PKELJlWbcWV5h_IEnQrKgkaFuQ85w0ASN2IM7IQHJRuZbMkj_dzVY58o_fzOPkqOzsifdqv6BATnf7Gavc7TjHQrXcpZhd3r9Rv7TPSHkP02YdnakNPE-YpJjv5GD6S08GOGS-P84I8Xf_8tbqt1g83d6sf68o1qpkqIVR5Z9CdQNajkG4YlGIWwfZdIzqwEh1KlOKrHlTbaMsc6J53reJCO-Tignw53N2l-Gcu-Wbrs8NxtAHjnA0XUoBiom2LtDlI9x1ywsHsUmmRXg0Ds8dpNuaA0-xxGmCm4Cy2z8eEudti_9_0j18RfD8IsPR88ZhMdh6Dw94ndJPpo38_4Q1p-oja</recordid><startdate>202005</startdate><enddate>202005</enddate><creator>Lee, Sehyung</creator><creator>Negishi, Makiko</creator><creator>Urakubo, Hidetoshi</creator><creator>Kasai, Haruo</creator><creator>Ishii, Shin</creator><general>Elsevier Ltd</general><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7X8</scope><orcidid>https://orcid.org/0000-0003-2327-9027</orcidid><orcidid>https://orcid.org/0000-0001-7137-4622</orcidid><orcidid>https://orcid.org/0000-0002-1816-0040</orcidid></search><sort><creationdate>202005</creationdate><title>Mu-net: Multi-scale U-net for two-photon microscopy image denoising and restoration</title><author>Lee, Sehyung ; Negishi, Makiko ; Urakubo, Hidetoshi ; Kasai, Haruo ; Ishii, Shin</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c474t-337684f8b3e1de36cff771ae0adb43b0a6ece6e6358f7948a1c08d2b97238ce23</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Deep learning</topic><topic>GAN</topic><topic>Image denoising</topic><topic>Two-photon microscopy image</topic><topic>U-net</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Lee, Sehyung</creatorcontrib><creatorcontrib>Negishi, Makiko</creatorcontrib><creatorcontrib>Urakubo, Hidetoshi</creatorcontrib><creatorcontrib>Kasai, Haruo</creatorcontrib><creatorcontrib>Ishii, Shin</creatorcontrib><collection>PubMed</collection><collection>CrossRef</collection><collection>MEDLINE - Academic</collection><jtitle>Neural networks</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Lee, Sehyung</au><au>Negishi, Makiko</au><au>Urakubo, Hidetoshi</au><au>Kasai, Haruo</au><au>Ishii, Shin</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Mu-net: Multi-scale U-net for two-photon microscopy image denoising and restoration</atitle><jtitle>Neural networks</jtitle><addtitle>Neural Netw</addtitle><date>2020-05</date><risdate>2020</risdate><volume>125</volume><spage>92</spage><epage>103</epage><pages>92-103</pages><issn>0893-6080</issn><eissn>1879-2782</eissn><abstract>Advances in two two-photon microscopy (2PM) have made three-dimensional (3D) neural imaging of deep cortical regions possible. However, 2PM often suffers from poor image quality because of various noise factors, including blur, white noise, and photo bleaching. In addition, the effectiveness of the existing image processing methods is limited because of the special features of 2PM images such as deeper tissue penetration but higher image noises owing to rapid laser scanning. To address the denoising problems in 2PM 3D images, we present a new algorithm based on deep convolutional neural networks (CNNs). The proposed model consists of multiple U-nets in which an individual U-net removes noises at different scales and then yields a performance improvement based on a coarse-to-fine strategy. Moreover, the constituent CNNs employ fully 3D convolution operations. Such an architecture enables the proposed model to facilitate end-to-end learning without any pre/post processing. Based on the experiments on 2PM image denoising, we observed that our new algorithm demonstrates substantial performance improvements over other baseline methods.</abstract><cop>United States</cop><pub>Elsevier Ltd</pub><pmid>32078964</pmid><doi>10.1016/j.neunet.2020.01.026</doi><tpages>12</tpages><orcidid>https://orcid.org/0000-0003-2327-9027</orcidid><orcidid>https://orcid.org/0000-0001-7137-4622</orcidid><orcidid>https://orcid.org/0000-0002-1816-0040</orcidid><oa>free_for_read</oa></addata></record> |
fulltext | fulltext |
identifier | ISSN: 0893-6080 |
ispartof | Neural networks, 2020-05, Vol.125, p.92-103 |
issn | 0893-6080 1879-2782 |
language | eng |
recordid | cdi_proquest_miscellaneous_2363071399 |
source | Access via ScienceDirect (Elsevier) |
subjects | Deep learning GAN Image denoising Two-photon microscopy image U-net |
title | Mu-net: Multi-scale U-net for two-photon microscopy image denoising and restoration |
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%3A38%3A34IST&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=Mu-net:%20Multi-scale%20U-net%20for%20two-photon%20microscopy%20image%20denoising%20and%20restoration&rft.jtitle=Neural%20networks&rft.au=Lee,%20Sehyung&rft.date=2020-05&rft.volume=125&rft.spage=92&rft.epage=103&rft.pages=92-103&rft.issn=0893-6080&rft.eissn=1879-2782&rft_id=info:doi/10.1016/j.neunet.2020.01.026&rft_dat=%3Cproquest_cross%3E2363071399%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=2363071399&rft_id=info:pmid/32078964&rft_els_id=S0893608020300368&rfr_iscdi=true |