Revealing architectural order with quantitative label-free imaging and deep learning.
deep learning
human
human tissue
inverse algorithms
label-free imaging
mouse
neuroscience
phase
physics of living systems
polarization
Journal
eLife
ISSN: 2050-084X
Titre abrégé: Elife
Pays: England
ID NLM: 101579614
Informations de publication
Date de publication:
27 07 2020
27 07 2020
Historique:
received:
27
01
2020
accepted:
24
07
2020
pubmed:
28
7
2020
medline:
13
2
2021
entrez:
28
7
2020
Statut:
epublish
Résumé
We report quantitative label-free imaging with phase and polarization (QLIPP) for simultaneous measurement of density, anisotropy, and orientation of structures in unlabeled live cells and tissue slices. We combine QLIPP with deep neural networks to predict fluorescence images of diverse cell and tissue structures. QLIPP images reveal anatomical regions and axon tract orientation in prenatal human brain tissue sections that are not visible using brightfield imaging. We report a variant of U-Net architecture, multi-channel 2.5D U-Net, for computationally efficient prediction of fluorescence images in three dimensions and over large fields of view. Further, we develop data normalization methods for accurate prediction of myelin distribution over large brain regions. We show that experimental defects in labeling the human tissue can be rescued with quantitative label-free imaging and neural network model. We anticipate that the proposed method will enable new studies of architectural order at spatial scales ranging from organelles to tissue. Microscopy is central to biological research and has enabled scientist to study the structure and dynamics of cells and their components within. Often, fluorescent dyes or trackers are used that can be detected under the microscope. However, this procedure can sometimes interfere with the biological processes being studied. Now, Guo, Yeh, Folkesson et al. have developed a new approach to examine structures within tissues and cells without the need for a fluorescent label. The technique, called QLIPP, uses the phase and polarization of the light passing through the sample to get information about its makeup. A computational model was used to decode the characteristics of the light and to provide information about the density and orientation of molecules in live cells and brain tissue samples of mice and human. This way, Guo et al. were able to reveal details that conventional microscopy would have missed. Then, a type of machine learning, known as ‘deep learning’, was used to translate the density and orientation images into fluorescence images, which enabled the researchers to predict specific structures in human brain tissue sections. QLIPP can be added as a module to a microscope and its software is available open source. Guo et al. hope that this approach can be used across many fields of biology, for example, to map the connectivity of nerve cells in the human brain or to identify how cells respond to infection. However, further work in automating other aspects, such as sample preparation and analysis, will be needed to realize the full benefits.
Autres résumés
Type: plain-language-summary
(eng)
Microscopy is central to biological research and has enabled scientist to study the structure and dynamics of cells and their components within. Often, fluorescent dyes or trackers are used that can be detected under the microscope. However, this procedure can sometimes interfere with the biological processes being studied. Now, Guo, Yeh, Folkesson et al. have developed a new approach to examine structures within tissues and cells without the need for a fluorescent label. The technique, called QLIPP, uses the phase and polarization of the light passing through the sample to get information about its makeup. A computational model was used to decode the characteristics of the light and to provide information about the density and orientation of molecules in live cells and brain tissue samples of mice and human. This way, Guo et al. were able to reveal details that conventional microscopy would have missed. Then, a type of machine learning, known as ‘deep learning’, was used to translate the density and orientation images into fluorescence images, which enabled the researchers to predict specific structures in human brain tissue sections. QLIPP can be added as a module to a microscope and its software is available open source. Guo et al. hope that this approach can be used across many fields of biology, for example, to map the connectivity of nerve cells in the human brain or to identify how cells respond to infection. However, further work in automating other aspects, such as sample preparation and analysis, will be needed to realize the full benefits.
Identifiants
pubmed: 32716843
doi: 10.7554/eLife.55502
pii: 55502
pmc: PMC7431134
doi:
pii:
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Subventions
Organisme : NICHD NIH HHS
ID : T32 HD007470
Pays : United States
Informations de copyright
© 2020, Guo et al.
Déclaration de conflit d'intérêts
SG, LY, JF, II, AK, MK, EH, DS, BC, NC, ML, MH, TN, SM No competing interests declared
Références
J Neurosci. 2011 Nov 9;31(45):16125-38
pubmed: 22072665
Light Sci Appl. 2019 Feb 6;8:23
pubmed: 30728961
J Opt. 2013 Sep 1;15(9):
pubmed: 24273640
Nat Biomed Eng. 2019 Jun;3(6):466-477
pubmed: 31142829
Front Neuroinform. 2011 Nov 11;5:28
pubmed: 22110430
Proc Natl Acad Sci U S A. 2016 Oct 18;113(42):E6352-E6361
pubmed: 27679846
J Microsc. 2008 Sep;231(3):419-32
pubmed: 18754996
Biophys J. 2000 Mar;78(3):1176-82
pubmed: 10692307
Opt Express. 2018 Dec 10;26(25):32888-32899
pubmed: 30645449
Opt Express. 2015 Oct 5;23(20):26672-82
pubmed: 26480180
Opt Express. 2017 Jul 10;25(14):15699-15712
pubmed: 28789083
J Cell Sci. 2014 Jul 15;127(Pt 14):2999-3004
pubmed: 25024457
Nat Methods. 2019 Dec;16(12):1233-1246
pubmed: 31133758
Biophys J. 2001 Mar;80(3):1557-67
pubmed: 11222316
Nature. 2007 Jan 11;445(7124):168-76
pubmed: 17151600
Opt Lett. 2006 Mar 15;31(6):775-7
pubmed: 16544620
Proc Natl Acad Sci U S A. 2012 Oct 9;109(41):16480-5
pubmed: 23012402
J Opt Soc Am A Opt Image Sci Vis. 2016 Jul 1;33(7):1396-408
pubmed: 27409699
Neuroimage. 2017 Aug 15;157:561-574
pubmed: 28602815
Cells. 2019 Oct 05;8(10):
pubmed: 31590363
Nat Methods. 2019 Dec;16(12):1215-1225
pubmed: 31285623
J Neurosci Methods. 2012 Aug 15;209(2):344-50
pubmed: 22743799
Front Neuroanat. 2018 Sep 27;12:75
pubmed: 30323745
J Struct Biol. 2011 Dec;176(3):302-6
pubmed: 21970947
Cereb Cortex. 2017 Mar 1;27(3):1779-1794
pubmed: 26874183
Bioarchitecture. 2014;4(6):189-202
pubmed: 26317264
Appl Opt. 2015 Nov 1;54(31):9213-27
pubmed: 26560576
Adv Exp Med Biol. 2021;1310:211-238
pubmed: 33834439
PLoS Comput Biol. 2016 Nov 4;12(11):e1005177
pubmed: 27814364
Science. 1955 Mar 11;121(3141):345-9
pubmed: 13237991
Appl Opt. 2015 Oct 1;54(28):8566-79
pubmed: 26479636
Biomed Opt Express. 2016 Sep 09;7(10):3940-3950
pubmed: 27867705
Nat Methods. 2012 Jan 15;9(3):255-8
pubmed: 22245809
J Microsc. 1995 Nov;180(Pt 2):140-7
pubmed: 8537959
Appl Opt. 1990 Sep 10;29(26):3810-5
pubmed: 20567488
Opt Express. 2010 Jun 7;18(12):12552-61
pubmed: 20588381
Cell. 2018 Apr 19;173(3):792-803.e19
pubmed: 29656897
Biophys J. 2011 Aug 17;101(4):985-94
pubmed: 21843491
Analyst. 2017 Apr 10;142(8):1207-1215
pubmed: 27840868
J R Soc Interface. 2015 Oct 6;12(111):20150734
pubmed: 26446561
Front Neuroinform. 2011 Dec 30;5:34
pubmed: 22232597
Nature. 2005 Feb 10;433(7026):597-603
pubmed: 15660108
Biomed Opt Express. 2017 Jun 07;8(7):3163-3197
pubmed: 28717561
Acta Histochem. 1980;66(1):32-9
pubmed: 6776776
Appl Opt. 2003 Jun 1;42(16):3009-17
pubmed: 12790452
Mol Biol Cell. 2017 Nov 7;28(23):3349-3359
pubmed: 28835378
J Microsc. 2018 Sep;271(3):315-324
pubmed: 29926918
Reprod Biomed Online. 2003 Jul-Aug;7(1):24-9
pubmed: 12930570
J Biomed Opt. 2008 Jan-Feb;13(1):014011
pubmed: 18315369
Elife. 2020 Dec 17;9:
pubmed: 33331817
PLoS Comput Biol. 2020 Sep 14;16(9):e1008193
pubmed: 32925919
Elife. 2020 Jul 27;9:
pubmed: 32716843
J Biophotonics. 2018 Aug;11(8):e201700356
pubmed: 29633576
Chromosoma. 1953;5(5):487-500
pubmed: 13082662
Nature. 2014 Apr 10;508(7495):207-14
pubmed: 24695228
Opt Lett. 2018 Feb 15;43(4):763-766
pubmed: 29443988
Nat Methods. 2018 Nov;15(11):917-920
pubmed: 30224672
Curr Opin Neurobiol. 2018 Jun;50:154-162
pubmed: 29579713
Brain Struct Funct. 2019 Jan;224(1):159-170
pubmed: 30293214
Neuroimage. 2011 Jan 15;54(2):1091-101
pubmed: 20832489
Neuroimage. 2018 Jan 15;165:56-68
pubmed: 29017866
Opt Express. 2015 May 4;23(9):11394-403
pubmed: 25969234
Dev Neurobiol. 2018 Feb;78(2):136-151
pubmed: 29082667
Front Aging Neurosci. 2015 Apr 14;7:48
pubmed: 25926791
Nature. 1952 Mar 1;169(4296):366-7
pubmed: 14919571
Front Neuroanat. 2009 Jun 01;3:5
pubmed: 19521542