Volume 51 Issue 11
Nov.  2022
Turn off MathJax
Article Contents

Yan Tianyu, He Ying, Wang Xinyu, Xu Xinyi, Xie Hui, Chen Xueli. Research progress on fast 3D fluorescence microscopic imaging (invited)[J]. Infrared and Laser Engineering, 2022, 51(11): 20220546. doi: 10.3788/IRLA20220546
Citation: Yan Tianyu, He Ying, Wang Xinyu, Xu Xinyi, Xie Hui, Chen Xueli. Research progress on fast 3D fluorescence microscopic imaging (invited)[J]. Infrared and Laser Engineering, 2022, 51(11): 20220546. doi: 10.3788/IRLA20220546

Research progress on fast 3D fluorescence microscopic imaging (invited)

doi: 10.3788/IRLA20220546
  • Received Date: 2022-08-03
  • Rev Recd Date: 2022-09-23
  • Publish Date: 2022-11-30
  • With the advantages of high resolution, high sensitivity, high molecule specificity and non-invasive, fluorescence microscopy can characterize the morphological and molecular functional information of samples at the micron or even nanometer scale, making it an important tool for life science research. As microbiology research continues to advance, fluorescence microscopy is expected to provide dynamic and 3D observation of microscopic biological structures and molecular events. This paper systematically reviews the research progress of fast 3D fluorescence microscopy in recent years, including the main technical means, improvement strategies and representative research results of point-scan imaging, wide-field imaging, and projection tomography to improve the imaging speed, expand the imaging dimension and enhance the imaging quality. In the end, we look forward to the future challenges and prospects of fast 3D fluorescence microscopic imaging techonology.
  • [1] Lakadamyali M. High resolution imaging of neuronal connec-tivity [J]. Journal of Microscopy, 2012, 248(2): 111-116. doi:  10.1111/j.1365-2818.2012.03638.x
    [2] Xu K, Babcock H P, Zhuang X W. Dual-objective STORM reveals three-dimensional filament organization in the actin cytoskeleton [J]. Nature Methods, 2012, 9(2): 185-188. doi:  10.1038/nmeth.1841
    [3] Jungmann R, Avendano M S, Woehrstein J B, et al. Multiplexed 3D cellular super-resolution imaging with DNA-PAINT and exchange-PAINT [J]. Nature Methods, 2014, 11(3): 313-318. doi:  10.1038/nmeth.2835
    [4] Lin J R, Fallahi-Sichani M, Sorger P K. Highly multiplexed imaging of single cells using a high-throughput cyclic immunofluorescence method [J]. Nature Communications, 2015, 6: 8390. doi:  10.1038/ncomms9390
    [5] Peng X H, Huang X S, Du K, et al. High spatiotemporal resolution and low photo-toxicity fluorescence imaging in live cells and in vivo [J]. Biochemical Society Transactions, 2019, 47(6): 1635-1650. doi:  10.1042/bst20190020
    [6] Kandel M E, He Y C R, Lee Y J, et al. Phase imaging with computational specificity (PICS) for measuring dry mass changes in sub-cellular compartments [J]. Nature Commu-nications, 2020, 11(1): 6256. doi:  10.1038/s41467-020-20062-x
    [7] Ghukasyan V V, Kao F J. Monitoring cellular metabolism with fluorescence lifetime of reduced nicotinamide adenine dinucleotide [J]. Journal of Physical Chemistry C, 2009, 113(27): 11532-11540. doi:  10.1021/jp810931u
    [8] Ouzounov D G, Wang T Y, Wang M R, et al. In vivo three-photon imaging of activity of GCaMP6-labeled neurons deep in intact mouse brain [J]. Nature Methods, 2017, 14(4): 388-390. doi:  10.1038/nmeth.4183
    [9] Zhang X D, Wang H X, Wang H, et al. Single-layered graphitic-C3 N4 quantum dots for two-photon fluorescence imaging of cellular nucleus [J]. Advanced Materials, 2014, 26(26): 4438-4443. doi:  10.1002/adma.201400111
    [10] von Diezmann A, Shechtman Y, Moerner W E. Three-dimensional localization of single molecules for super resolution imaging and single-particle tracking [J]. Chemical Reviews, 2017, 117(11): 7244-7275. doi:  10.1021/acs.chemrev.6b00629
    [11] Hedde P N, Ranjit S, Gratton E. 3D fluorescence anisotropy imaging using selective plane illumination microscopy [J]. Optics Express, 2015, 23(17): 22308-22317. doi:  10.1364/oe.23.022308
    [12] Wu Y, Wu X D, Lu R, et al. Resonant scanning with large field of view reduces photobleaching and enhances fluorescence yield in STED microscopy [J]. Scientific Reports, 2015, 5: 14766. doi:  10.1038/srep14766
    [13] Palero J, Santos S, Artigas D, et al. A simple scanless two-photon fluorescence microscope using selective plane illumination [J]. Optics Express, 2010, 18(8): 8491-8498. doi:  10.1364/oe.18.008491
    [14] Zhang Z K, Cong L, Bai L, et al. Light-field microscopy for fast volumetric brain imaging [J]. Journal of Neuroscience Methods, 2021, 352: 109083. doi:  10.1016/j.jneumeth.2021.109083
    [15] Zhu L, Zhang W, Elnatan D, et al. Faster STORM using compressed sensing [J]. Nature Methods, 2012, 9(7): 721-723. doi:  10.1038/nmeth.1978
    [16] Soini J T, Schrader M, Hanninen P E, et al. Image formation and data acquisition in a stage scanning 4Pi confocal fluorescence microscope [J]. Applied Optics, 1997, 36(34): 8929-8934. doi:  10.1364/ao.36.008929
    [17] van Munster E B, Goedhart J, Kremers G J, et al. Combination of a spinning disc confocal unit with frequency-domain fluorescence lifetime imaging microscopy [J]. Cytometry Part A, 2007, 71A(4): 207-214. doi:  10.1002/cyto.a.20379
    [18] Conchello J A, Lichtman J W. Optical sectioning microscopy [J]. Nature Methods, 2005, 2(12): 920-931. doi:  10.1038/nmeth815
    [19] Qi X L, Yang T, Li L H, et al. Fluorescence micro-optical sectioning tomography using acousto-optical deflector-based confocal scheme [J]. Neurophotonics, 2015, 2(4): 041406. doi:  10.1117/1.NPh.2.4.041406
    [20] Zong W J, Wu R L, Chen S Y, et al. Miniature two-photon microscopy for enlarged field-of-view, multi-plane and long-term brain imaging [J]. Nature Methods, 2021, 18(1): 46-49. doi:  10.1038/s41592-020-01024-z
    [21] Ruprecht A K, Wiesendanger T F, Tiziani H J. Chromatic confocal microscopy with a finite pinhole size [J]. Optics Letters, 2004, 29(18): 2130-2132. doi:  10.1364/ol.29.002130
    [22] Ishii H, Otomo K, Takahashi T, et al. Focusing new light on brain functions: multiphoton microscopy for deep and super-resolution imaging [J]. Neuroscience Research, 2022, 179: 24-30. doi:  10.1016/j.neures.2021.11.011
    [23] Arbabi E, Li J Q, Hutchins R J, et al. Two-photon microscopy with a double-wavelength metasurface objectivel ens [J]. Nano Letters, 2018, 18(8): 4943-4948. doi:  10.1021/acs.nanolett.8b01737
    [24] Zhu S J, Yang Q L, Antaris A L, et al. Molecular imaging of biological systems with a clickable dye in the broad 800-to 1, 700-nm near-infrared window [J]. Proceedings of the National Academy of Sciences of the United States of America, 2017, 114(5): 962-967. doi:  10.1073/pnas.1617990114
    [25] Yang W J, Carrillo-Reid L, Bando Y, et al. Simultaneous two-photon imaging and two-photon optogenetics of cortical circuits in three dimensions [J]. Elife, 2018, 7: e32671. doi:  10.7554/eLife.32671
    [26] Scully A D, Ostler R B, MacRobert A J, et al. Laser line-scanning confocal fluorescence imaging of the photodynamic action of aluminum and zinc phthalocyanines in V79-4 Chinese hamster fibroblasts [J]. Photochemistry and Photobiology, 1998, 68(2): 199-204. doi:  10.1111/j.1751-1097.1998.tb02489.x
    [27] Piyawattanametha W, Barretto R P J, Ko T H, et al. Fast-scanning two-photon fluorescence imaging based on a microelectromechanical systems two-dimensional scanning mirror [J]. Optics Letters, 2006, 31(13): 2018-2020. doi:  10.1364/ol.31.002018
    [28] Boutilier R M, Park J S, Lee H. High-speed two-photon laser scanning microscopy imaging of in vivo blood cells in rapid circulation at velocities of up to 1.2 millimeters per second [J]. Current Optics and Photonics, 2018, 2(6): 595-605. doi:  10.3807/copp.2018.2.6.595
    [29] Zhang T, Hernandez O, Chrapkiewicz R, et al. Kilohertz two-photon brain imaging in awake mice [J]. Nature Methods, 2019, 16(11): 1119-1122. doi:  10.1038/s41592-019-0597-2
    [30] Woods E, Courtney J, Scholz D, et al. Tracking protein dynamics with photoconvertible Dendra2 on spinning disk confocal systems [J]. Journal of Microscopy, 2014, 256(3): 197-207. doi:  10.1111/jmi.12172
    [31] Oketani R, Suda H, Uegaki K, et al. Visible-wavelength two-photon excitation microscopy with multifocus scanning for volumetric live-cell imaging [J]. Journal of Biomedical Optics, 2020, 25(1): 014502. doi:  10.1117/1.Jbo.25.1.014502
    [32] Chien Y F, Lin J Y, Yeh P T, et al. Dual GRIN lens two-photon endoscopy for high-speed volumetric and deep brain imaging [J]. Biomedical Optics Express, 2021, 12(1): 162-172. doi:  10.1364/boe.405738
    [33] York A G, Parekh S H, Nogare D D, et al. Resolution doubling in live, multicellular organisms via multifocal structured illumination microscopy [J]. Nature Methods, 2012, 9(7): 749-754. doi:  10.1038/nmeth.2025
    [34] Chen Z Y, Mc Larney B, Rebling J, et al. High-speed large-field multifocal illumination fluorescence microscopy [J]. Laser & Photonics Reviews, 2020, 14(2): 1900070. doi:  10.1002/lpor.201900070
    [35] Wu J L, Liang Y J, Chen S, et al. Kilohertz two-photon fluorescence microscopy imaging of neural activity in vivo [J]. Nature Methods, 2020, 17(3): 287-290. doi:  10.1038/s41592-020-0762-7
    [36] Duocastella M, Sun B, Arnold C B. Simultaneous imaging of multiple focal planes for three-dimensional microscopy using ultra-high-speed adaptive optics [J]. Journal of Biomedical Optics, 2012, 17(5): 050505. doi:  10.1117/1.Jbo.17.5.050505
    [37] Weisenburger S, Tejera F, Demas J, et al. Volumetric Ca2+ imaging in the mouse brain using hybrid multiplexed sculpted light microscopy [J]. Cell, 2019, 177(4): 1050-1066. doi:  10.1016/j.cell.2019.03.011
    [38] Lu R W, Liang Y J, Meng G H, et al. Rapid mesoscale volumetric imaging of neural activity with synaptic resolution [J]. Nature Methods, 2020, 17(3): 291-294. doi:  10.1038/s41592-020-0760-9
    [39] Hao X, Li Y M, Fu S, et al. Review of 4Pi fluorescence nanoscopy [J]. Engineering, 2022, 11: 146-153. doi:  10.1016/j.eng.2020.07.028
    [40] Tortarolo G, Sun Y S, Teng K W, et al. Photon-separation to enhance the spatial resolution of pulsed STED microscopy [J]. Nanoscale, 2019, 11(4): 1754-1761. doi:  10.1039/c8nr07485b
    [41] Velasco M G M, Zhang M Y, Antonello J, et al. 3D super-resolution deep-tissue imaging in living mice [J]. Optica, 2021, 8(4): 442-450. doi:  10.1364/optica.416841
    [42] Li B, Wu C Y, Wang M R, et al. An adaptive excitation source for high-speed multiphoton microscopy [J]. Nature Methods, 2020, 17(2): 163-166. doi:  10.1038/s41592-019-0663-9
    [43] Hillman E M C, Voleti V, Li W Z, et al. Light-sheet microscopy in neuroscience[M]//Annual Review of Neuro-science, 2019.
    [44] Gibbs H C, Mota S M, Hart N A, et al. Navigating the light-sheet image analysis software landscape: concepts for driving cohesion from data acquisition to analysis [J]. Frontiers in Cell and Developmental Biology, 2021, 9: 739079. doi:  10.3389/fcell.2021.739079
    [45] Poola P K, Afzal M I, Yoo Y, et al. Light sheet microscopy for histopathology applications [J]. Biomedical Engineering Letters, 2019, 9(3): 279-291. doi:  10.1007/s13534-019-00122-y
    [46] Gu P C, Huang Z X, Ping M, et al. Thinner and longer working distance light sheet illumination and microscopic imaging [J]. IEEE Journal of Selected Topics in Quantum Electronics, 2021, 27(4): 7300107. doi:  10.1109/jstqe.2020.2996606
    [47] Liu T L, Upadhyayula S, Milkie D E, et al. Observing the cell in its native state: Imaging subcellular dynamics in multicellular organisms [J]. Science, 2018, 360(6386): 284-284. doi:  10.1126/science.aaq1392
    [48] Fei P, Nie J, Lee J, et al. Subvoxel light-sheet microscopy for high-resolution high-throughput volumetric imaging of large biomedical specimens [J]. Advanced Photonics, 2019, 1(1): 016002. doi:  10.1117/1.AP.1.1.016002
    [49] Fahrbach F O, Voigt F F, Schmid B, et al. Rapid 3D light-sheet microscopy with a tunable lens [J]. Optics Express, 2013, 21(18): 21010-21026. doi:  10.1364/oe.21.021010
    [50] Haslehurst P, Yang Z Y, Dholakia K, et al. Fast volume-scanning light sheet microscopy reveals transient neuronal events [J]. Biomedical Optics Express, 2018, 9(5): 2154-2167. doi:  10.1364/boe.9.002154
    [51] Lin P Y, Hwang S P L, Lee C H, et al. Two-photon scanned light sheet fluorescence microscopy with axicon imaging for fast volumetric imaging [J]. Journal of Biomedical Optics, 2021, 26(11): 116503. doi:  10.1117/1.Jbo.26.11.116503
    [52] Olarte O E, Andilla J, Artigas D, et al. Decoupled illumination detection in light sheet microscopy for fast volumetric imaging [J]. Optica, 2015, 2(8): 702-705. doi:  10.1364/optica.2.000702
    [53] Yang B, Chen X Y, Wang Y N, et al. Epi-illumination SPIM for volumetric imaging with high spatial-temporal resolution [J]. Nature Methods, 2019, 16(6): 501-504. doi:  10.1038/s41592-019-0401-3
    [54] Yang B, Lange M, Millett-Sikking A, et al. DaXi-high-resolution, large imaging volume and multi-view single-objective light-sheet microscopy [J]. Nature Methods, 2022, 19(4): 461-469. doi:  10.1038/s41592-022-01417-2
    [55] Cai Y H, Chen Y Z, Xia Y Q, et al. Single-lens light-sheet fluorescence microscopy based on micro-mirror array [J]. Laser & Photonics Reviews, 2022, 16(8): 2100026. doi:  10.1002/lpor.202100026
    [56] Wang D P, Zhu Z J, Xu Z Y, et al. Neuroimaging with light field microscopy: a mini review of imaging systems [J]. European Physical Journal-Special Topics, 2022, 231(4): 749-761. doi:  10.1140/epjs/s11734-021-00367-8
    [57] Broxton M, Grosenick L, Yang S, et al. Wave optics theory and 3-D deconvolution for the light field microscope [J]. Optics Express, 2013, 21(21): 25418-25439. doi:  10.1364/oe.21.025418
    [58] Wang D P, Roy S, Rudzite A M, et al. High-resolution light-field microscopy with patterned illumination [J]. Biomedical Optics Express, 2021, 12(7): 3887-3901. doi:  10.1364/boe.425742
    [59] Prevedel R, Yoon Y G, Hoffmann M, et al. Simultaneous whole-animal 3D imaging of neuronal activity using light-field microscopy [J]. Nature Methods, 2014, 11(7): 727-730. doi:  10.1038/nmeth.2964
    [60] Wagner N, Norlin N, Gierten J, et al. Instantaneous isotropic volumetric imaging of fast biological processes [J]. Nature Methods, 2019, 16(6): 497-500. doi:  10.1038/s41592-019-0393-z
    [61] Wu J M, Lu Z, Jian D, et al. Iterative tomography with digital adaptive optics permits hour-long intravital observation of 3D subcellular dynamics at millisecond scale [J]. Cell, 2021, 184(12): 3318-3332.e17. doi:  10.1016/j.cell.2021.04.029
    [62] Pan Z, Lu M, Xia S. Diffraction-assisted light field microscopy for microtomography and digital volume correlation with improved spatial resolution [J]. Experimental Mechanics, 2019, 59(5): 713-724. doi:  10.1007/s11340-019-00522-2
    [63] He K, Wang X L, Wang Z H W, et al. Snapshot multifocal light field microscopy [J]. Optics Express, 2020, 28(8): 12108-12120. doi:  10.1364/oe.390719
    [64] Geng Q, Fu Z Q, Chen S C. High-resolution 3D light-field imaging [J]. Journal of Biomedical Optics, 2020, 25(10): 106502. doi:  10.1117/1.Jbo.25.10.106502
    [65] Wang Z Q, Zhu L X, Zhang H, et al. Real-time volumetric reconstruction of biological dynamics with light-field microscopy and deep learning [J]. Nature Methods, 2021, 18(5): 551-556. doi:  10.1038/s41592-021-01058-x
    [66] Liu J D, Xu T F, Yue W R, et al. Light-field moment microscopy with noise reduction [J]. Optics Express, 2015, 23(22): 29154-29162. doi:  10.1364/oe.23.029154
    [67] Wang H C, Chen N, Zheng S S, et al. Fast and high-resolution light field acquisition using defocus modulation [J]. Applied Optics, 2018, 57(1): A250-A256. doi:  10.1364/ao.57.00a250
    [68] Truong T V, Holland D B, Madaan S, et al. High-contrast, synchronous volumetric imaging with selective volume illumination microscopy [J]. Communications Biology, 2020, 3(1): 74. doi:  10.1038/s42003-020-0787-6
    [69] Wang D P, Xu S, Pant P, et al. Hybrid light-sheet and light-field microscope for high resolution and large volume neuroimaging [J]. Biomedical Optics Express, 2019, 10(12): 6595-6610. doi:  10.1364/boe.10.006595
    [70] Zhang Z K, Bai L, Cong L, et al. Imaging volumetric dynamics at high speed in mouse and zebrafish brain with confocal light field microscopy [J]. Nature Biotechnology, 2021, 39(1): 74-83. doi:  10.1038/s41587-020-0628-7
    [71] Schneckenburger H, Richter V. Laser scanning versus wide-field-choosing the appropriate microscope in life sciences [J]. Applied Sciences-Basel, 2021, 11(2): 733. doi:  10.3390/app11020733
    [72] Luo J T, Li C K, Liu Q L, et al. Super-resolution structured illumination microscopy reconstruction using a least-squares solver [J]. Frontiers in Physics, 2020, 8: 118. doi:  10.3389/fphy.2020.00118
    [73] Leung B O, Chou K C. Review of super-resolution fluorescence microscopy for biology [J]. Applied Spectroscopy, 2011, 65(9): 967-980. doi:  10.1366/11-06398
    [74] Kurdzialek S, Demkowicz-Dobrzanski R. Super-resolution optical fluctuation imaging-fundamental estimation theory perspective [J]. Journal of Optics, 2021, 23(7): 075701. doi:  10.1088/2040-8986/ac059c
    [75] Prakash K, Diederich B, Reichelt S, et al. Super-resolution structured illumination microscopy: past, present and future [J]. Philosophical Transactions of the Royal Society A-Mathematical Physical and Engineering Sciences, 2021, 379(2199): 20200143. doi:  10.1098/rsta.2020.0143
    [76] Gustafsson M G L, Shao L, Carlton P M, et al. Three-dimensional resolution doubling in wide-field fluorescence microscopy by structured illumination [J]. Biophysical Journal, 2008, 94(12): 4957-4970. doi:  10.1529/biophysj.107.120345
    [77] Fiolka R, Beck M, Stemmer A. Structured illumination in total internal reflection fluorescence microscopy using a spatial light modulator [J]. Optics Letters, 2008, 33(14): 1629-1631. doi:  10.1364/ol.33.001629
    [78] Schwertner M, Booth M J, Wilson T. Specimen-induced distortions in light microscopy [J]. Journal of Microscopy, 2007, 228(1): 97-102. doi:  10.1111/j.1365-2818.2007.01827.x
    [79] Lin R Z, Kipreos E T, Zhu J, et al. Subcellular three-dimensional imaging deep through multicellular thick samples by structured illumination microscopy and adaptive optics [J]. Nature Communications, 2021, 12(1): 3148. doi:  10.1038/s41467-021-23449-6
    [80] Jones S A, Shim S H, He J, et al. Fast, three-dimensional super-resolution imaging of live cells [J]. Nature Methods, 2011, 8(6): 499-505. doi:  10.1038/nmeth.1605
    [81] Stein S C, Huss A, Hahnel D, et al. Fourier interpolation stochastic optical fluctuation imaging [J]. Optics Express, 2015, 23(12): 16154-16163. doi:  10.1364/oe.23.016154
    [82] Samanta S, Gong W J, Li W, et al. Organic fluorescent probes for stochastic optical reconstruction microscopy (STORM): Recent highlights [J]. Coordination Chemistry Reviews, 2019, 380: 17-34. doi:  10.1016/j.ccr.2018.08.006
    [83] Lelek M, Gyparaki M, Melina T, et al. Single-molecule localization microscopy [J]. Nature Reviews Methods Primers, 2021, 1: 39. doi:  10.1038/s43586-021-00038-x
    [84] Huang B, Wang W Q, Bates M, et al. Three-dimensional super-resolution imaging by stochastic optical reconstruction microscopy [J]. Science, 2008, 319(5864): 810-813. doi:  10.1126/science.1153529
    [85] Wang Y H, Jia S, Zhang H F, et al. Blind sparse inpainting reveals cytoskeletal filaments with sub-Nyquist localization [J]. Optica, 2017, 4(10): 1277-1284. doi:  10.1364/optica.4.001277
    [86] Zeng Z P, Chen X Z, Wang H N, et al. Fast super-resolution imaging with ultra-high labeling density achieved by joint tagging super-resolution optical fluctuation imaging [J]. Scientific Reports, 2015, 5: 8359. doi:  10.1038/srep08359
    [87] GruSsmayer K S, Geissbuehler S, Descloux A, et al. Spectral cross-cumulants for multicolor super-resolved SOFI imaging [J]. Nature Communications, 2020, 11(1): 3023. doi:  10.1038/s41467-020-16841-1
    [88] Chen X Z, Zeng Z P, Li R Q, et al. Superior performance with sCMOS over EMCCD in super-resolution optical fluctuation imaging [J]. Journal of Biomedical Optics, 2016, 21(6): 066007. doi:  10.1117/1.Jbo.21.6.066007
    [89] Sharpe J. Optical projection tomography [J]. Annual Review of Biomedical Engineering, 2004, 6: 209-228. doi:  10.1146/annurev.bioeng.6.040803.140210
    [90] Birk U J, Rieckher M, Konstantinides N, et al. Correction for specimen movement and rotation errors for in-vivo optical projection tomography [J]. Biomedical Optics Express, 2010, 1(1): 87-96. doi:  10.1364/boe.1.000087
    [91] Vinegoni C, Fexon L, Feruglio P F, et al. High throughput transmission optical projection tomography using low cost graphics processing unit [J]. Optics Express, 2009, 17(25): 22320-22332. doi:  10.1364/oe.17.022320
    [92] Bassi A, Fieramonti L, D'Andrea C, et al. In vivo label-free three-dimensional imaging of zebrafish vasculature with optical projection tomography [J]. Journal of Biomedical Optics, 2011, 16(10): 100502. doi:  10.1117/1.3640808
    [93] Arranz A, Dong D, Zhu S P, et al. In-vivo optical tomography of small scattering specimens: time-lapse 3D imaging of the head eversion process in Drosophila melanogaster [J]. Scientific Reports, 2014, 4: 7325. doi:  10.1038/srep07325
    [94] McGinty J, Taylor H B, Chen L, et al. In vivo fluorescence lifetime optical projection tomography [J]. Biomedical Optics Express, 2011, 2(5): 1340-1350. doi:  10.1364/boe.2.001340
    [95] Juntunen C, Woller I M, Sung Y J. Hyperspectral three-dimensional fluorescence imaging using snapshot Optical Tomography [J]. Sensors, 2021, 21(11): 3652. doi:  10.3390/s21113652
    [96] Sharpe J, Ahlgren U, Perry P, et al. Optical projection tomography as a tool for 3D microscopy and gene expression studies [J]. Science, 2002, 296(5567): 541-545. doi:  10.1126/science.1068206
    [97] Zhu S P, Dong D, Birk U J, et al. Automated motion correction for in vivo optical projection tomography [J]. IEEE Transactions on Medical Imaging, 2012, 31(7): 1358-1371. doi:  10.1109/tmi.2012.2188836
    [98] Cheddad A, Svensson C, Sharpe J, et al. Image processing assisted algorithms for optical projection tomography [J]. IEEE Transactions on Medical Imaging, 2012, 31(1): 1-15. doi:  10.1109/tmi.2011.2161590
    [99] Chen L L, McGinty J, Taylor H B, et al. Incorporation of an experimentally determined MTF for spatial frequency filtering and deconvolution during optical projection tomography reconstruction [J]. Optics Express, 2012, 20(7): 7323-7337. doi:  10.1364/oe.20.007323
    [100] Gong C C, Zeng L, Wang C X. Image reconstruction model for limited-angle CT based on prior image induced relative total variation [J]. Applied Mathematical Modelling, 2019, 74: 586-605. doi:  10.1016/j.apm.2019.05.020
    [101] Gong C C, Zeng L. Anisotropic structure property based image reconstruction method for limited-angle computed tomography [J]. Journal of X-Ray Science and Technology, 2021, 29(6): 1079-1102. doi:  10.3233/xst-210954
    [102] Chen Z Q, Jin X, Li L, et al. A limited-angle CT reconstruction method based on anisotropic TV minimization [J]. Physics in Medicine and Biology, 2013, 58(7): 2119-2141. doi:  10.1088/0031-9155/58/7/2119
    [103] Wang N, Chen D F, Chen D, et al. Feasibility study of limited-angle reconstruction for in vivo optical projection tomography based on novel sample fixation [J]. IEEE Access, 2019, 7: 87681-87691. doi:  10.1109/access.2019.2925096
    [104] Chen X L, Zhu S P, Wang H Y, et al. Accelerated stimulated Raman projection tomography by sparse reconstruction from sparse-view data [J]. IEEE Transactions on Biomedical Engineering, 2020, 67(5): 1293-1302. doi:  10.1109/tbme.2019.2935301
    [105] Wang H Y, Wang N, Xie H, et al. Two-stage deep learning network-based few-view image reconstruction for parallel-beam projection tomography [J]. Quantitative Imaging in Medicine and Surgery, 2022, 12(4): 2535-2551. doi:  10.21037/qims-21-778
    [106] Zhong Q Y, Li A A, Jin R, et al. High-definition imaging using line-illumination modulation microscopy [J]. Nature Methods, 2021, 18(3): 309-315. doi:  10.1038/s41592-021-01074-x
    [107] Zhu S J, Herraiz S, Yue J Y, et al. 3D NIR-II molecular imaging distinguishes targeted organs with high-performance NIR-II bioconjugates [J]. Advanced Materials, 2018, 30(13): 1705799. doi:  10.1002/adma.201705799
    [108] Rodriguez C, Chen A, Rivera J A, et al. An adaptive optics module for deep tissue multiphoton imaging in vivo [J]. Nature Methods, 2021, 18(10): 1259-1264. doi:  10.1038/s41592-021-01279-0
    [109] Fan J T, Suo J L, Wu J M, et al. Video-rate imaging of biological dynamics at centimetre scale and micrometre resolution [J]. Nature Photonics, 2019, 13(11): 809-816. doi:  10.1038/s41566-019-0474-7
    [110] Ozeki Y, Umemura W, Otsuka Y, et al. High-speed molecular spectral imaging of tissue with stimulated Raman scattering [J]. Nature Photonics, 2012, 6(12): 844-850. doi:  10.1038/nphoton.2012.263
    [111] Chen X L, Zhang C, Lin P, et al. Volumetric chemical imaging by stimulated Raman projection microscopy and tomography [J]. Nature Communications, 2017, 8: 15117. doi:  10.1038/ncomms15117
  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(7)

Article Metrics

Article views(260) PDF downloads(85) Cited by()

Related
Proportional views

Research progress on fast 3D fluorescence microscopic imaging (invited)

doi: 10.3788/IRLA20220546
  • 1. School of Life Science and Technology, Xidian University, Xi’an 710126, China
  • 2. Xi’an Key Laboratory of Intelligent Sensing and Regulation of trans-Scale Life Information, Xidian University, Xi’an 710126, China

Abstract: With the advantages of high resolution, high sensitivity, high molecule specificity and non-invasive, fluorescence microscopy can characterize the morphological and molecular functional information of samples at the micron or even nanometer scale, making it an important tool for life science research. As microbiology research continues to advance, fluorescence microscopy is expected to provide dynamic and 3D observation of microscopic biological structures and molecular events. This paper systematically reviews the research progress of fast 3D fluorescence microscopy in recent years, including the main technical means, improvement strategies and representative research results of point-scan imaging, wide-field imaging, and projection tomography to improve the imaging speed, expand the imaging dimension and enhance the imaging quality. In the end, we look forward to the future challenges and prospects of fast 3D fluorescence microscopic imaging techonology.

    • 三维生物微观信息的可视化对于细胞代谢、组织微结构、脑神经科学等生命科学领域的研究有着重要意义[1-3]。荧光显微成像技术具有样本损伤小、分辨率高、背景低、通量大以及分子特异性强的优势[4-6],可以对细胞代谢、神经元活动中的分子事件进行高特异性追踪[7-9]。通过轴向步进扫描等方式,荧光显微成像技术可以采集样本内不同深度的图像并进行重建,实现三维成像,进而更全面地揭示微观生命活动的原理和机制[10]。但是由于数据维度的增加,相比于二维平面成像,三维荧光显微成像的整体数据量将膨胀两三个数量级,导致成像的实时性大幅下降[11]。为了解决这一问题,快速扫描策略[12]、选择性照明方式 [13]、光场成像[14]以及稀疏化采样[15]等方法被提出并不断优化,使得快速三维荧光显微成像成为可能。文中将系统性阐述和总结各种类型的快速三维荧光显微成像的技术手段,并展望其未来的发展方向与前景。

    • 使用具有层析能力的荧光显微镜对样本进行不同深度的逐层成像获得图像堆栈,是实现三维荧光显微成像的常用方式[16]。传统的反射式宽场照明显微镜进行成像时,其焦外荧光信号会进入探测器成为背景噪声,降低图像对比度并导致较差的层析能力[17],即使成像速度很快,在不配合逐层切片这种对样本具有破坏性的技术时,也很难实现高质量的三维图像重建。另一方面,和传统的反射式宽场照明荧光显微镜相比,以激光扫描共聚焦显微镜(confocal laser scanning microscope, CLSM)和双光子显微镜(two-photon microscope, TPM)为代表的扫描式显微镜具有更高的穿透深度、成像信噪比和光学切片能力[18-20],为三维图像的光学重建提供了有力的数据采集工具。典型的CLSM系统结构如图1(a)所示。CLSM和TPM均使用了逐点照明和逐点检测的成像策略来抑制临近像素之间的信号干扰,其差异在于:CLSM通过像方的针孔滤除焦平面的荧光信号,实现了成像信噪比、横向分辨率以及光学切片能力的提升[21];TPM则是利用高瞬时功率的脉冲激光在物镜焦点处能够产生足够高的光子密度使荧光分子同时吸收两个长波长光子的现象,仅在聚焦光斑位置进行选择性激发,实现了极低的成像背景以及优良的层析能力[22],并且因为激发波长较长,组织散射效应减少而具备了更深的成像深度[23]

      Figure 1.  Schematic diagram of the point-scanning system and its typical application results. (a) Structural diagram of confocal laser scanning microscope; (b) Multicolor confocal images and 3D reconstruction of brain tissue section[24]; (c) Single-photon and two-photon volumetric imaging of brain neurons, and imaging depths with different illumination strategies[22]

      CLSM和TPM由于其良好的图像质量已被广泛应用于如细胞成像、在体组织成像、脑神经元研究(图1(b)~(c))等领域中。2017年,Zhu等人使用自行合成的近红外二区造影剂,在其自行搭建的CLSM系统中实现了对厚度为150 μm的脑组织切片的三维体积成像[24]。2018年,Yang等人将双光子光遗传学与双光子钙体积成像相结合,以单细胞级的分辨率进行了在体三维测量和操纵小鼠大脑皮层的神经活动[25]。然而,由于逐点扫描与逐点检测的成像策略,点扫描式成像若以轴向像素数接近横向像素数为目标,获取可以用于三维重建的图像堆栈是一个缺乏实时性的过程,解决这一问题的主要方法是提高扫描速度,将数据通量提高至每秒数千万乃至上亿像素。

      减少一个扫描维度提升成像速度的线扫描策略是早期被提出的方案[26],但是只从一个方向屏蔽焦外信号会降低成像质量,提高机械扫描组件扫描转子的频率获得更高的扫描速度也是一种常见的思路。早在2007年,Piyawattanametha等人开发的二维单晶硅扫描镜便已用于TPM和显微内窥镜中,其快轴采集速率可达3.52 kHz[27]。2019年,Boutilier等人通过将72面多面镜安装在每分钟55000转的电机上以提高TPM水平方向上的快轴扫描速度,能够以90 帧/s的速度捕获512 pixel×512 pixel的图像[28]

      进一步,提高扫描转子的利用效率也是提高扫描速度的主要方向之一。基于传统的机械式扫描方案中,通过偏转单个光束的传播方向获得运动的单个焦点进行扫描的策略,发展出了同时偏转复数个光束进行多焦点并行式高速扫描的方法(图2(a))。2019年,Zhang等人使用扫描振镜和微透镜阵列,实现了400个焦点的并行式扫描,并能以1000帧/s速率对清醒小鼠的大脑进行深达300 μm的双光子成像[29]。此外,多焦点扫描策略允许不同于传统的光束偏转式扫描策略,其中具有代表性的是转盘式扫描显微镜,其结构如图2(b)所示。利用包含针孔阵列的转盘,通过转盘高速转动对样本进行多焦点扫描的成像帧率可以比单焦点扫描提升一两个数量级[30]。2021年,Oketani等人使用针孔阵列盘作为扫描单元,在可见光波段通过双光子成像重建了宫颈癌细胞中高尔基体的三维结构,并对其在三维空间上的运动进行了实时监测[31],其体积成像速率可达2 s/帧(图2(c))。

      Figure 2.  Realization of fast scanning technology and its typical application. (a) Multifocal scanning device; (b) Spinning disk multifocal scanning; (c) Fast scanning technique for real-time observation of 3D movement of the Golgi apparatus[31]; (d) Time delayed multifocal scanning; (e) Axial scanning with tunable lens; (f) 3D functional imaging of mouse neurons through a tunable lens[32], 180 µm× 180 µm× 165 µm, temporal resolution: 0.25 s; (g) 3D imaging result of a zebrafish embryo through multifocal structured illumination microscope. Red box indicates a dividing cell[33]

      然而,多焦点的扫描策略通常要求面阵探测器同步探测多焦点激发的荧光信号,对相机的最快曝光时间与光子探测效率等性能要求非常严苛[34]。基于时间延迟的多路复用的多焦点扫描技术可以在不同焦点间产生照明延时,进而允许高速高灵敏度的单点探测器(如光电倍增管)采集多焦点的时间信号序列进行图像重构。2020年,Wu等人在TPM中,令非平行的脉冲激光束在两个近乎平行的反射镜之间进行多次反射以分离不同波矢方向的子脉冲,进而在样本面获得了数十个具有时间延迟的焦点,配合一维扫描振镜与光电倍增管实现了近乎线扫描的采样速度(主要结构见图2(d)),最快能以3000帧/s的速度获得二维图像[35]

      但是,仅提高横向扫描速率不能完全解决体积成像的实时性问题,还需提升轴向扫描的效率。传统的基于位移台的轴向扫描在惯性控制、位移精度和位移距离等方面存在限制,难以实现快速轴向扫描[36]。将变焦透镜引入成像系统通过高速变焦进行轴向快速扫描成为了一种新的思路。变焦透镜执行轴向扫描的原理如图2(e)所示。2021年,Chien将轴向扫描频率高达1 MHz的可调谐声学梯度指数镜头应用于双光子显微镜中,实现了以亚秒级的体积速率对小鼠大脑钙动力学信息的实时监控(图2(f)) [32]。进一步,Weisenburger通过在照明光路中设置一个可以沿光轴方向快速移动的反射镜也实现了轴向快速扫描,并结合多路复用技术实现了17 Hz的体积成像速率[37]。此外,利用贝塞尔光等无衍射光束在不损失横向分辨率的前提下获得轴向长聚焦光斑也是一种实现快速体积成像的手段[38]

      另一方面,充分利用点扫描照明策略的优势,进一步增强空间分辨率,实现超分辨三维成像也是目前提高数据通量的一个发展方向。例如,York等人开发了一种多焦点结构光照明显微镜,利用数字微镜进行多焦点并行扫描,并通过对采集的数据施加数字针孔提升成像质量,最终获得了145 nm的横向分辨率(图2(g)),突破了传统的光学分辨率极限[33];使用两个反向相对的物镜获得近乎全方向孔径角的4Pi显微镜技术,可以实现轴向分辨率高达7倍的提升[39];经典的受激辐射损耗显微镜可以通过照明光束和损耗光束的叠加获得远小于衍射极限的激发光斑,配合三维扫描可以实现超分辨三维成像[40]。2021年,Velasco等人在受激辐射损耗显微镜的基础上,结合双光子照明、自适应光学和近红外荧光探针等技术,获得了深度可达164 μm的三维超分辨荧光显微图像[41]

      总之,通过多焦点、静态扫描装置和变焦透镜等进行点扫描式的高速三维荧光显微成像是可行的。但是在提高扫描速度的同时,单个采样点的光子产量及曝光时间会不可避免地下降,这对光子预算的规划要求更为严格[42],并且系统的复杂化会导致成本的提高。这些问题限制了点扫描式高速三维荧光显微成像的普及。另外,超分辨技术虽然和点扫描式成像有良好的兼容性,但应当注意权衡分辨率、视野和成像时间之间的关系。

    • 除了点扫描策略外,在轴向进行选择性照明也是一种抑制离焦信号的思路,而光片显微镜就是其中具有代表性的一种。光片显微镜采用了薄片状的照明光源替代了常规显微镜中的柱形光源,由于只有在光片内部的荧光分子被高效率激发,成像时只会产生较少的焦外信号,提高了图像对比度[43-45],其系统结构如图3(a)所示。和点扫描式成像相比,光片显微镜由于宽场照明策略,大幅度降低了对生物样本的光毒性[46],并且在横向上可以由面阵探测器直接成像,采集图像堆栈只需执行一个维度的扫描,非常适合长时间实时成像的应用。2018年,Liu 等人将晶格光片显微镜与自适应光学相结合,实现了对大体积多细胞样本中亚细胞过程的无创无像差成像[47]。2019年,Fei等人将像素超分辨技术加入到光片显微镜中,实现了以非拼接的形式对大体积样本的各向同性的高分辨率成像[48]

      光片显微镜同样需要考虑轴向扫描速率的问题,而基于振镜和可变焦透镜的高速扫描策略同样可用于光片显微镜中。Fahrbach等人利用透镜扫描实现了以510 帧/s的速度对跳动的斑马鱼心脏内的17个平面进行成像[49],相当于每秒进行30次体积扫描(图3(b))。Haslehurst等人在光片显微镜中加入了振镜进行快速轴向扫描,并配合电控可调谐镜头同步采集图像,实现了对哺乳动物脑组织中神经元树突状突起的快速功能性钙成像[50]

      Figure 3.  Structure of light-sheet microscope and typical volumetric imaging result. (a) Structure of traditional light-sheet microscope; (b) Volumetric imaging result of zebrafish heart[49], including 2D image sequence (left) and 3D reconstruction (right), scale bar: 50 μm; (c) Time-lapse images of a zebrafish heart[51], the line chart on the right is the heart beat dynamics, scale bar: 30 μm

      另一方面,通过扩展景深的方法也是提高光片显微镜成像速度的一个方向。Lin等人通过将轴棱锥加入到双光子光片荧光显微镜中,产生了传统镜头20倍以上的拓展景深[51],进而以最高30 Hz的体积采集速率捕获了斑马鱼幼虫中动态跳动的心脏和红细胞的图像(图3(c))。Olarte等人通过使用波前编码技术在光片显微镜中扩展检测光学器件景深的方法也实现了快速体积成像。这种技术允许在扩展轴向范围内的任何位置调节光片,他们的结果表明该方案可在不移动样本的条件下进行每秒高达70次的体积成像[52]

      此外,光片显微镜的照明臂和成像臂的正交结构限制了样本空间,并且影响了短工作距离的大数值孔径物镜的适用,因此单物镜式光片显微镜被开发出来。2019年,Yang等人在落射式荧光显微镜中生成了倾斜的光片进行选择性照明,并在像方光路选择了区别于照明光路的视野,实现了在物方仅需单个物镜的光片显微镜[53],并进一步在2022年通过优化设计突破其原本在成像视野方面的限制[54]。Cai等人在落射式荧光显微镜中生成了沿轴向传播的光片照明,又在像方沿光轴方向设置了微镜阵列以模拟轴向共焦狭缝阵列去分离不同焦平面的信号,配合一维扫描实现了单物镜、正入射式的光片显微镜[55]

    • 选择性照明显微镜进行体积成像时由于其扫描装置通常是必需的,因此在成像速度和简化系统结构方面依然受到制约。光场显微镜通过在像方光路中设置微透镜阵列的方式同时获取光的强度和角度信息,通过算法可以在不执行扫描的条件下重建样本的三维信息,具有天然的速度优势,是一种具有大视野、大景深的快速三维荧光显微成像技术[56-57],其光学结构如图4(a)所示。

      Figure 4.  Structure of light-field microscope and typical volumetric imaging result. (a) Structure of traditional light-field microscope; (b) Typical reconstruction results at different depths of the worm brain[59], scale bar: 50 µm; (c) Blood flow imaging at 200 Hz volumetric imaging rate[60], 200 µm×200 µm×200 µm

      但是,光场显微镜的三维成像能力本质上是以牺牲横向分辨率为代价的[58],需要在横向和轴向分辨率之间做出权衡,并且其在三维视场上的分辨率并不均匀,这些缺陷限制了光场显微镜的实际应用。因此对光场显微镜的分辨率提升是目前其主要发展方向。早在2014年,Prevedel等人就通过反卷积的图像重建方法获得了1.4 μm的有效分辨率,并实现了对整个蠕虫和斑马鱼幼体大脑以单神经元级的精度在三维视野内同时进行的神经元活动进行功能成像(图4(b)),体积成像速率达到20帧/s[59]。2021年,Wu 等人提出了一个称为数字自适应光学扫描光场相互迭代断层扫描的计算成像框架,将光场显微镜的横向分辨率和轴向分辨率分别提升至了220 nm和400 nm[61]。此外,在光路中加入衍射光学元件也是增强分辨率的有效手段。2019年,Pan等人通过在样品和显微镜物镜之间插入透射式光栅获得不同衍射阶数的光场图像并推导出系统的点扩散函数以用于图像重建,在不影响横向分辨率的情况下提升了轴向分辨率[62]。2020年,He等人通过在光场显微镜的傅里叶平面上放置衍射光学元件实现了一种快照多焦光场显微成像方法,该方法可以同时收集多个焦移光场,并将所有异焦光场一起反卷积,可以在不牺牲横向分辨率及其均匀性的条件下获得较大的场深[63]

      另一方面,系统改良和重建策略的优化可以使光场显微镜的成像速度得到进一步提高。2019年,Wagner等人使用两个互相垂直的物镜同时获取正交光场,进行双视图数据融合并反卷积,实现了高达200 Hz的无运动伪影体积成像(图4(c))[60]。2020年,Geng等人利用高分辨率的二维图像和低分辨率的四维光场图像,通过反卷积和相位恢复的混合算法将图像重建的计算速度提高了4倍[64]。2021年,Wang等人结合光场显微镜与深度学习成像技术的方案实现了以单细胞级的分辨率对跳动的斑马鱼心脏中的血流进行成像,其体积成像速率也达到了200 Hz[65]

      此外,光场显微镜存在和传统宽场反射式荧光显微镜同样的问题,即景深外被激发的荧光会成为干扰信号,在成像对比度、信噪比和空间分辨率方面都有所欠缺[66-67]。但另一方面,由于光场显微镜的系统结构简单,具有良好的兼容性,可以和其他先进照明策略进行结合以提升图像质量。2020年,Truong等人构建了一个基于光场的选择性体积照明显微镜,其照明范围被控制在感兴趣的体积内,有效降低了光场成像的背景噪声,并获得了35%的分辨率增强效果[68]。类似地,2020年,Wang等人将光片照明技术应用于光场显微镜,使其成像对比度增强了10倍,成像信噪比提升了3.2倍[69]。Zhang等人将共焦检测方案与光场显微镜结合,选择性地从照明光内部收集荧光信号,获取了光学切片能力以及最高可达600 μm的成像深度[70]

    • 围绕宽场显微镜开发超分辨成像技术,可以在保持较低的光毒性的同时获得更丰富的细节[71],同样是目前三维荧光显微成像技术的一个重要发展方向。基于宽场显微镜实现超分辨成像的方法较多,大致可以分为结构光照明超分辨成像[72]、单分子定位超分辨成像[73]和光学涨落超分辨成像[74]三类。

      结构光照明超分辨成像技术通过对样本施加干涉条纹并相移来实现样本图像的超分辨率重建,可以将空间分辨率提高一倍。其中,二维结构光照明技术只需两束光干涉以及三次相位变换,但只能提升横向分辨率[75];而三维结构光则需要三束光产生干涉条纹并进行15次相位变换(其典型系统结构如图5(a)所示),并可以重建出样本的三维超分辨率荧光图像(图5(b)) [76],且使用空间光调制器可以将相位变换过程的耗时控制在很短的时间内[77]。然而,三维结构光照明技术对于像差非常敏感,因而在生物组织内成像深度仅为数个微米,通常仅能覆盖单个细胞的厚度[78]。针对这一问题,Lin等人在2021年通过将自适应光学技术引入到三维结构光照明成像中以消除像差,成功将其成像深度提高至80 μm[79]

      Figure 5.  Structure of 3D-SIM and typical imaging result of super resolution technology. (a) Representative system structure; (b) Comparison of imaging results of fluorescent microspheres with conventional microscope (left) and 3D-SIM (right)[76]; (c) Dynamic 3D super resolution imaging results of transferrin clusters[80], scale bar: 50 nm; (d) Imaging results of rat hippocampal neuron[81], including widefield imaging (left), classical super resolution optical fluctuation reconstruction (middle), and super resolution optical fluctuation reconstruction using Fourier interpolation (right)

      单分子定位超分辨成像技术通过对视野内的荧光分子进行稀疏性激发,使临近的荧光分子不同时发光,进而允许在每次成像时对零散的荧光分子以纳米级精度定位,这样的成像进行成千上万次即可重构出一张超分辨图像[82-83]。基于单分子定位的成像技术已被证明可以在横向和轴向均获得超分辨率能力。早在2008年,庄小威等人就提出了一种三维单分子定位超分辨成像方案,该方案使用了一个柱面镜在像平面生成椭圆光斑,通过椭圆长短轴的变化来表征荧光在轴向的位置,最终实现了横向30 nm、轴向60 nm的单分子定位精度[84],并在2011年以类似的分辨率实现了0.5 Hz的体积成像速率(图5(c)) [80]。为了进一步降低单分子定位超分辨成像技术对数据量的要求,Wang等人于2017年提出了一种适用于单分子定位的盲修复图像重建方法,可以从低密度图像中恢复原本无法解析的精细结构,对原始数据的需求量仅为传统方法的百分之一,因而可以将成像速度提高两个数量级[85]

      光学涨落超分辨成像技术通过捕捉独立分布的荧光分子的闪烁状态而实现分辨率增强[86]。具体而言,视野内被激光照射的荧光分子会随时间随机闪烁,因此可以通过采集一组时间图像序列,依据图像同一位置的像素值变化来分离临近的荧光分子,进而重构出超分辨率图像[87]。光学涨落超分辨成像技术可以直接使用宽场荧光显微镜采集的时间图像序列完成超分辨图像重建,具有低成本、适用广泛的优点,但是其分辨率提升效果受限于探测器像素尺寸[88]。对此,Stein等人曾提出一种基于傅里叶变换的插值方案,可以突破像素尺寸的限制,重建出像素密度更高的图像且不会引入伪影(图5(d)),提高了光学涨落超分辨技术对分辨率增强的上限[81]

    • 使用了各类新技术的宽场式三维荧光显微成像在克服了传统宽场显微镜缺陷的同时,凭借更低的光毒性以及数据采集维度的优势,成为了实时体积成像的强有力工具,允许科研人员对微观视野下短暂的分子事件进行动态追踪。而超分辨技术的加入则使得宽场显微成像可以在时间分辨率和空间分辨率之间进行取舍。

    • 类似于使用X射线的计算机断层扫描,在光学波段对样本进行多角度弹道式照明获取投影数据集可以对样本实现三维断层重建,即光学投影断层成像技术(optical projection tomography, OPT) [89-91]。典型OPT系统结构如图6(a)所示。OPT已被证明可以用于某些光学透明度较高的活体生物进行全身断层重建。Bassi等人利用OPT技术以无标记的形式实现了对弱散射活体样本的血管网络的可视化与三维断层重建[92]。Arranz等人在没有进行光学透明化处理的条件下,利用OPT技术对黑腹果蝇蛹进行了三维断层扫描[93],并捕捉到了蛹内头部外翻过程的体积图像(图6(b))。此外,结合荧光标记技术,OPT可对生物样本内部的光吸收和荧光信号进行三维追踪。McGinty等人使用绿色荧光蛋白标记了斑马鱼胚胎后,成功对其进了包含荧光寿命信息的光学投影断层扫描[94]

      Figure 6.  OPT system and typical 3D reconstruction results. (a) Traditional OPT system structure; (b) Fluorescent 3D reconstruction results of Drosophila melanogaster pupae[93]

      进一步,在缩小成像视野并提高放大倍率后,OPT技术同样支持以微米级的空间分辨率对生物样本在微观尺度上执行断层重建。Juntunen等人于2021年通过结合高光谱技术实现了对直径约10 μm的向日葵花粉粒的快速光学断层重建,获得了高质量的光吸收与荧光图像[95]。总得来说,相比于点扫描需要对样本进行数百万至数千万次采样才能重建出三维图像,结合OPT技术的三维荧光显微成像仅需获得数百个不同角度的投影即可完成采样并断层重建,并且可以检测厚度在毫米量级的样本[96]

      减少数据采集时间是进一步提高OPT技术成像速度的直接且高效的手段。首先,OPT成像对于投影数据的角度信息的精确度要求较高[97],而对角度的精准定位和校正是一个相对耗时的过程。 Cheddad等人针对这一问题开发出了样品在旋转轴上的半自动精确定位方法,并结合数据采集后的自动校正算法,获得了0.01 pixel的校正精度,可以减少数据采集时间并提高重建后的图像质量[98]。另外,提高重建时对数据的利用效率能够有效控制对原始数据量的需求,进而提高成像速度。Chen等人采用了基于调制传递函数的频率截止滤波器作为重建时的附加滤波器,在抑制重建伪影的同时降低了对数据量的要求,将数据采集速度提高了4倍[99]

      此外,经典的OPT重建方法要求对样本进行圆周式、高频次采样,这仍是一个耗时相对较长的过程,并且采集全角度投影的方法很难用于对大尺寸样本的局部观测。因此,开发适用于有限角度投影以及稀疏投影的重建方法可以显著提高OPT技术的成像速度及其实用性。对于有限角度投影重建,由于采样角度不足,滤波反投影(filtered back projection, FBP)和代数重建法等经典方法难以重建出良好的图像[100-101]。Chen等人受压缩感知理论启发,提出了一种名为“基于各项异性全变分极小化”的重建方法,可以在重建有限角度的投影数据时获得明显优于经典方法的图像质量[102]。Wang等人在2019年通过优化样本固定方式提高了活体样本的纵向检测能力,并利用代数重建与先验信息相结合的方法,最终仅需采集130°范围内的投影即可重建出质量可接受的图像(图7(a))[103]。在稀疏投影重建方面,Chen等人在2020年提出了一种基于稀疏视图数据的稀疏重建方案,结合代数重建和全变分正则化的迭代重建方法,用于重建的投影数量可以降低至15个[104],比滤波反投影重建方法所需的投影数量少了一个数量级。Wang等人在2022年将深度学习应用于三维重建中,提出了一种基于两阶段深度学习网络的框架(Two-stage deep learning net-work, TSDLN),用于稀疏投影的重建与去噪[105],可以将所需的投影数量进一步降低至9个(图7(b))。

      Figure 7.  Reconstruction results under limited angles or sparse sampling. (a) Results of 180° and 130° tomographic reconstruction[103]; (b) Reconstruction results of filtered back-projection using 180 projection data(left) and reconstruction results of 2-stage deep learning network using 9 projection data[105]

      OPT显微成像技术可以对厚样本内部的荧光信号和光吸收进行三维断层分析,有限角度重建和稀疏采样技术的加入可以降低其对样本几何形状及尺寸的限制,减少采样时间,扩大了该技术的检测目标的范围。但同时,OPT技术对样本的光学透过率要求较高,对于不透明的样本需要先进行透明化预处理,因此可以检测的活体样本的种类有限,并会丢失在透明化处理中被去除的生物信息。

    • 笔者归纳和分析了相对常见的三维荧光显微成像方案及其实现高速成像的技术策略。可以认为,现有的技术方案在一定程度上已经允许对部分微观尺度上的生物分子事件进行三维动态跟踪,但是仍存在一些因素在限制实时体积成像的进一步发展。首先,由于动态三维成像产生的庞大数据通量以及某些重建方法对算力的高要求,导致数据处理的实时性较差,该问题的解决可能需要结合数据压缩技术及计算机科学的进步[106];其次,由于生物组织对光的散射以及自体荧光背景,光学显微成像的深度被限制在数百微米内,近年来兴起的在近红外二区(900~1700 nm)窗口的荧光显微成像因为更低的散射和自体荧光提供了深层成像的可能[107],进一步,随着成像深度的加深,像差的增加会降低图像质量,而自适应光学可以有效缓解这一问题[108];第三,传统显微成像技术中有限的视场限制了数据吞吐量的上限,这一问题可以通过改进物镜设计而有效改善[109]。最后,由于荧光团的种类有限以及同步激发的困难,基于荧光标记的显微成像技术可能难以描述样本的全部分子事件,而拉曼光谱成像技术由于其特有的分子间化学键指纹表征能力[110-111],可以提供更加丰富的光学信息。笔者相信,随着各种先进光学成像技术的发展与融合,实时三维光学显微成像将会有进一步突破,提供更加完善的微观生物信息。

Reference (111)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return