Real-time image denoising of mixed Poisson–Gaussian noise in fluorescence microscopy images using ImageJ

Fluorescence microscopy imaging speed is fundamentally limited by the measurement signal-to-noise ratio (SNR). To improve image SNR for a given image acquisition rate, computational denoising techniques can be used to suppress noise. However, common techniques to estimate a denoised image from a sin...

Full description

Saved in:
Bibliographic Details
Published in:Optica 2022-03, Vol.9 (4)
Main Authors: Mannam, Varun, Zhang, Yide, Zhu, Yinhao, Nichols, Evan, Wang, Qingfei, Sundaresan, Vignesh, Zhang, Siyuan, Smith, Cody, Bohn, Paul W., Howard, Scott S.
Format: Article
Language:English
Subjects:
OTHER INSTRUMENTATION
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page
container_issue 4
container_start_page
container_title Optica
container_volume 9
creator Mannam, Varun
Zhang, Yide
Zhu, Yinhao
Nichols, Evan
Wang, Qingfei
Sundaresan, Vignesh
Zhang, Siyuan
Smith, Cody
Bohn, Paul W.
Howard, Scott S.
description Fluorescence microscopy imaging speed is fundamentally limited by the measurement signal-to-noise ratio (SNR). To improve image SNR for a given image acquisition rate, computational denoising techniques can be used to suppress noise. However, common techniques to estimate a denoised image from a single frame either are computationally expensive or rely on simple noise statistical models. These models assume Poisson or Gaussian noise statistics, which are not appropriate for many fluorescence microscopy applications that contain quantum shot noise and electronic Johnson–Nyquist noise, therefore a mixture of Poisson and Gaussian noise. In this paper, we show convolutional neural networks (CNNs) trained on mixed Poisson and Gaussian noise images to overcome the limitations of existing image denoising methods. The trained CNN is presented as an open-source ImageJ plugin that performs real-time image denoising (within tens of milliseconds) with superior performance (SNR improvement) compared to conventional fluorescence microscopy denoising methods. The method is validated on external datasets with out-of-distribution noise, contrast, structure, and imaging modalities from the training data and consistently achieves high-performance ( > 8 d B ) denoising in less time than other fluorescence microscopy denoising methods.
format article
fullrecord <record><control><sourceid>osti</sourceid><recordid>TN_cdi_osti_scitechconnect_1880378</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>1880378</sourcerecordid><originalsourceid>FETCH-osti_scitechconnect_18803783</originalsourceid><addsrcrecordid>eNqNjM0KgkAUhYcoSMp3GNoLY6Nl6-h3FdFehvFqV3QmvAq16x16w54krRYtW50fvnN6zJlKGXjTUM76P37IXKJcCOHLQIQL4bD8CKrwaiyBY6ky4AkYi4Qm4zblJV4h4Ye2IGue98dGNUSoDO-YdmF4WjS2AtJgNLS4rixpe7l9zog376ddF_ZjNkhVQeB-dcQm69VpufUs1RiTxhr0WVtjQNexH0VCziP5F_QC1yNLyA</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Real-time image denoising of mixed Poisson–Gaussian noise in fluorescence microscopy images using ImageJ</title><source>EZB Electronic Journals Library</source><creator>Mannam, Varun ; Zhang, Yide ; Zhu, Yinhao ; Nichols, Evan ; Wang, Qingfei ; Sundaresan, Vignesh ; Zhang, Siyuan ; Smith, Cody ; Bohn, Paul W. ; Howard, Scott S.</creator><creatorcontrib>Mannam, Varun ; Zhang, Yide ; Zhu, Yinhao ; Nichols, Evan ; Wang, Qingfei ; Sundaresan, Vignesh ; Zhang, Siyuan ; Smith, Cody ; Bohn, Paul W. ; Howard, Scott S. ; Univ. of Notre Dame, IN (United States)</creatorcontrib><description>Fluorescence microscopy imaging speed is fundamentally limited by the measurement signal-to-noise ratio (SNR). To improve image SNR for a given image acquisition rate, computational denoising techniques can be used to suppress noise. However, common techniques to estimate a denoised image from a single frame either are computationally expensive or rely on simple noise statistical models. These models assume Poisson or Gaussian noise statistics, which are not appropriate for many fluorescence microscopy applications that contain quantum shot noise and electronic Johnson–Nyquist noise, therefore a mixture of Poisson and Gaussian noise. In this paper, we show convolutional neural networks (CNNs) trained on mixed Poisson and Gaussian noise images to overcome the limitations of existing image denoising methods. The trained CNN is presented as an open-source ImageJ plugin that performs real-time image denoising (within tens of milliseconds) with superior performance (SNR improvement) compared to conventional fluorescence microscopy denoising methods. The method is validated on external datasets with out-of-distribution noise, contrast, structure, and imaging modalities from the training data and consistently achieves high-performance ( &gt; 8 d B ) denoising in less time than other fluorescence microscopy denoising methods.</description><identifier>ISSN: 2334-2536</identifier><identifier>EISSN: 2334-2536</identifier><language>eng</language><publisher>United States: Optical Society of America</publisher><subject>OTHER INSTRUMENTATION</subject><ispartof>Optica, 2022-03, Vol.9 (4)</ispartof><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><orcidid>0000000218666092 ; 0000000298311514 ; 0000000332466799 ; 0000000294633970</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>230,315,786,790,891</link.rule.ids><backlink>$$Uhttps://www.osti.gov/biblio/1880378$$D View this record in Osti.gov$$Hfree_for_read</backlink></links><search><creatorcontrib>Mannam, Varun</creatorcontrib><creatorcontrib>Zhang, Yide</creatorcontrib><creatorcontrib>Zhu, Yinhao</creatorcontrib><creatorcontrib>Nichols, Evan</creatorcontrib><creatorcontrib>Wang, Qingfei</creatorcontrib><creatorcontrib>Sundaresan, Vignesh</creatorcontrib><creatorcontrib>Zhang, Siyuan</creatorcontrib><creatorcontrib>Smith, Cody</creatorcontrib><creatorcontrib>Bohn, Paul W.</creatorcontrib><creatorcontrib>Howard, Scott S.</creatorcontrib><creatorcontrib>Univ. of Notre Dame, IN (United States)</creatorcontrib><title>Real-time image denoising of mixed Poisson–Gaussian noise in fluorescence microscopy images using ImageJ</title><title>Optica</title><description>Fluorescence microscopy imaging speed is fundamentally limited by the measurement signal-to-noise ratio (SNR). To improve image SNR for a given image acquisition rate, computational denoising techniques can be used to suppress noise. However, common techniques to estimate a denoised image from a single frame either are computationally expensive or rely on simple noise statistical models. These models assume Poisson or Gaussian noise statistics, which are not appropriate for many fluorescence microscopy applications that contain quantum shot noise and electronic Johnson–Nyquist noise, therefore a mixture of Poisson and Gaussian noise. In this paper, we show convolutional neural networks (CNNs) trained on mixed Poisson and Gaussian noise images to overcome the limitations of existing image denoising methods. The trained CNN is presented as an open-source ImageJ plugin that performs real-time image denoising (within tens of milliseconds) with superior performance (SNR improvement) compared to conventional fluorescence microscopy denoising methods. The method is validated on external datasets with out-of-distribution noise, contrast, structure, and imaging modalities from the training data and consistently achieves high-performance ( &gt; 8 d B ) denoising in less time than other fluorescence microscopy denoising methods.</description><subject>OTHER INSTRUMENTATION</subject><issn>2334-2536</issn><issn>2334-2536</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><recordid>eNqNjM0KgkAUhYcoSMp3GNoLY6Nl6-h3FdFehvFqV3QmvAq16x16w54krRYtW50fvnN6zJlKGXjTUM76P37IXKJcCOHLQIQL4bD8CKrwaiyBY6ky4AkYi4Qm4zblJV4h4Ye2IGue98dGNUSoDO-YdmF4WjS2AtJgNLS4rixpe7l9zog376ddF_ZjNkhVQeB-dcQm69VpufUs1RiTxhr0WVtjQNexH0VCziP5F_QC1yNLyA</recordid><startdate>20220322</startdate><enddate>20220322</enddate><creator>Mannam, Varun</creator><creator>Zhang, Yide</creator><creator>Zhu, Yinhao</creator><creator>Nichols, Evan</creator><creator>Wang, Qingfei</creator><creator>Sundaresan, Vignesh</creator><creator>Zhang, Siyuan</creator><creator>Smith, Cody</creator><creator>Bohn, Paul W.</creator><creator>Howard, Scott S.</creator><general>Optical Society of America</general><scope>OTOTI</scope><orcidid>https://orcid.org/0000000218666092</orcidid><orcidid>https://orcid.org/0000000298311514</orcidid><orcidid>https://orcid.org/0000000332466799</orcidid><orcidid>https://orcid.org/0000000294633970</orcidid></search><sort><creationdate>20220322</creationdate><title>Real-time image denoising of mixed Poisson–Gaussian noise in fluorescence microscopy images using ImageJ</title><author>Mannam, Varun ; Zhang, Yide ; Zhu, Yinhao ; Nichols, Evan ; Wang, Qingfei ; Sundaresan, Vignesh ; Zhang, Siyuan ; Smith, Cody ; Bohn, Paul W. ; Howard, Scott S.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-osti_scitechconnect_18803783</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>OTHER INSTRUMENTATION</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Mannam, Varun</creatorcontrib><creatorcontrib>Zhang, Yide</creatorcontrib><creatorcontrib>Zhu, Yinhao</creatorcontrib><creatorcontrib>Nichols, Evan</creatorcontrib><creatorcontrib>Wang, Qingfei</creatorcontrib><creatorcontrib>Sundaresan, Vignesh</creatorcontrib><creatorcontrib>Zhang, Siyuan</creatorcontrib><creatorcontrib>Smith, Cody</creatorcontrib><creatorcontrib>Bohn, Paul W.</creatorcontrib><creatorcontrib>Howard, Scott S.</creatorcontrib><creatorcontrib>Univ. of Notre Dame, IN (United States)</creatorcontrib><collection>OSTI.GOV</collection><jtitle>Optica</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Mannam, Varun</au><au>Zhang, Yide</au><au>Zhu, Yinhao</au><au>Nichols, Evan</au><au>Wang, Qingfei</au><au>Sundaresan, Vignesh</au><au>Zhang, Siyuan</au><au>Smith, Cody</au><au>Bohn, Paul W.</au><au>Howard, Scott S.</au><aucorp>Univ. of Notre Dame, IN (United States)</aucorp><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Real-time image denoising of mixed Poisson–Gaussian noise in fluorescence microscopy images using ImageJ</atitle><jtitle>Optica</jtitle><date>2022-03-22</date><risdate>2022</risdate><volume>9</volume><issue>4</issue><issn>2334-2536</issn><eissn>2334-2536</eissn><notes>USDOE Office of Science (SC), Biological and Environmental Research (BER)</notes><notes>0019312; SC0019312</notes><abstract>Fluorescence microscopy imaging speed is fundamentally limited by the measurement signal-to-noise ratio (SNR). To improve image SNR for a given image acquisition rate, computational denoising techniques can be used to suppress noise. However, common techniques to estimate a denoised image from a single frame either are computationally expensive or rely on simple noise statistical models. These models assume Poisson or Gaussian noise statistics, which are not appropriate for many fluorescence microscopy applications that contain quantum shot noise and electronic Johnson–Nyquist noise, therefore a mixture of Poisson and Gaussian noise. In this paper, we show convolutional neural networks (CNNs) trained on mixed Poisson and Gaussian noise images to overcome the limitations of existing image denoising methods. The trained CNN is presented as an open-source ImageJ plugin that performs real-time image denoising (within tens of milliseconds) with superior performance (SNR improvement) compared to conventional fluorescence microscopy denoising methods. The method is validated on external datasets with out-of-distribution noise, contrast, structure, and imaging modalities from the training data and consistently achieves high-performance ( &gt; 8 d B ) denoising in less time than other fluorescence microscopy denoising methods.</abstract><cop>United States</cop><pub>Optical Society of America</pub><orcidid>https://orcid.org/0000000218666092</orcidid><orcidid>https://orcid.org/0000000298311514</orcidid><orcidid>https://orcid.org/0000000332466799</orcidid><orcidid>https://orcid.org/0000000294633970</orcidid></addata></record>
fulltext fulltext
identifier ISSN: 2334-2536
ispartof Optica, 2022-03, Vol.9 (4)
issn 2334-2536
2334-2536
language eng
recordid cdi_osti_scitechconnect_1880378
source EZB Electronic Journals Library
subjects OTHER INSTRUMENTATION
title Real-time image denoising of mixed Poisson–Gaussian noise in fluorescence microscopy images using ImageJ
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-09-21T11%3A35%3A40IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-osti&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Real-time%20image%20denoising%20of%20mixed%20Poisson%E2%80%93Gaussian%20noise%20in%20fluorescence%20microscopy%20images%20using%20ImageJ&rft.jtitle=Optica&rft.au=Mannam,%20Varun&rft.aucorp=Univ.%20of%20Notre%20Dame,%20IN%20(United%20States)&rft.date=2022-03-22&rft.volume=9&rft.issue=4&rft.issn=2334-2536&rft.eissn=2334-2536&rft_id=info:doi/&rft_dat=%3Costi%3E1880378%3C/osti%3E%3Cgrp_id%3Ecdi_FETCH-osti_scitechconnect_18803783%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true