Open Access
Add to BibSonomy
Abstract
Three-dimensional (3D) fluorescence imaging is important to accurately capture and understand biological structures and phenomena. However, because of its slow acquisition speed, it was difficult to implement 3D fluorescence imaging for imaging flow cytometry. Especially, modern flow cytometers operate at a flow velocity of 1–10 m/s, and no 3D fluorescence imaging technique was able to capture cells at such high velocity. Here, we present a high-speed 3D fluorescence imaging technique in which a set of optical cross sections of a cell is captured within a single frame of a camera by combining strobe light-sheet excitation and optofluidic spatial transformation. Using this technique, we demonstrated 3D fluorescence imaging of cells flowing at a velocity of over 10 m/s, which is the fastest to our knowledge. Such technology can allow integration of 3D imaging with flow systems of common flow cytometers and cell sorters.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Three-dimensional (3D) fluorescence imaging becomes increasingly important in various fields of life sciences for obtaining accurate structural information of intra- and inter-cellular events and systems [1,2]. For obtaining 3D fluorescence images of cells, imaging techniques such as the conventional confocal scanning microscopy and the more recently developed light-sheet microscopy (LSM) are popularly utilized. Especially, LSM has gathered attention with the ability to rapidly obtain high-content, large-volume 3D images of biological specimens such as embryos and brains [1,3–5].
On the other hand, imaging of fast flowing cells with imaging flow cytometry (IFC) becomes of increasing interest in fields such as hematology, immunology, drug screening, and clinical diagnosis [6–9]. IFC is a recently emerging technique that combines the morphological characterization ability of microscopy with the high-throughput single-cell characterization ability of flow cytometry [7,8,10]. IFC pursues fast image sampling of individual cells with a diameter of typically around 10 µm while successively aligning them in a single flow stream at a high velocity, typically 1–10 m/s in commercial flow cytometers [11]. For the last decades, IFC technologies adopting 2D imaging methods have advanced significantly and achieved cell imaging at such flow velocities utilizing various high-speed imaging techniques [12–16]. However, 2D imaging fundamentally lacks z-axis spatial resolution, resulting in loss of spatial information, occlusion of objects, and blurring by focal depth. Therefore, in cases where 3D spatial information is critical such as analyzing cell organelle and protein localization, IFCs based on 3D imaging methods are expected to become essential.
Still, 3D imaging of fast-moving objects remains too slow to be implemented in the high-speed IFC because the extra dimension in 3D imaging increases the time required for image acquisition compared to 2D imaging. In the LSM methods, a fundamental bottle neck is that the imaging speed throughput is limited by the rate of optical sectioning defined by the frame rate of pixel arrayed cameras [17–20]. Even with a very high-speed volumetric imaging such as SCAPE [21,22], the frame rate was not fast enough to capture the 3D cell images in a flow faster than 20 mm/sec at a micrometer resolution. In analogous to confocal microscopes, a camera-less point scanning method was able to obtain 3D images of cells whereas the reported flow velocity was 0.2 m/s [23], which was not as fast as conventional flow cytometers working at 1–10 m/s.
Here, we present a high-speed 3D IFC method utilizing strobe light-sheet imaging with optofluidic spatial transformation which we call FLITS (optoFluidic Light-sheet Imaging with Transformation of Spatial information). By performing optofluidic scanning of each cell through optical pulse trains, FLITS enables spatial mapping of cell cross sections onto different positions of the camera with fluorescence light-sheet imaging. As a result, it is able to surpass the speed limit of light-sheet microscopy which was limited by the camera frame rate, and we were able to capture 3D images of single cells at a flow velocity of over 10 m/s, which is more than an order faster than existing reports.
2. Principle and methods
2.1 Principle of FLITS
FLITS is performed using a strobe light-sheet excitation that is slightly angled to the flowing channel as in Fig. 1. As a cell passes through this diagonal light sheet, each strobe pulse excite a cross section of a cell at different time points. Because the light sheet is slightly angled to the flowing channel, different cross sections of the cell are excited at these different time points. When this is viewed from the direction orthogonal to the light sheet, in which the camera is placed, different cross sections of the cell excited at different time points appear at different regions of the light sheet. Consequently, by imaging these cross sections with a single camera exposure, all cross sections of a 3D cell image can be captured within a single camera frame.
Fig. 1. Schematic of FLITS. A cell is allowed to flow through a strobe light-sheet illumination created by a pulsed light source. The light sheet is oriented with a slight angle to the direction of the flow. As the cell passes through this illumination, different cross sections of the cell are illuminated at different time points. These cross sections are acquired with a single exposure of a camera, while each cross section is imaged onto a different position of the camera sensor.
The imaging speed of FLITS, or the velocity of the cell that it can capture, is governed by the repetition rate of the optical pulse. For a cell velocity of 1 to 10 m/s, a repetition rate of several tens to several hundreds of kilohertz is sufficient. For example, for a cell with a size of 10 µm in diameter and flowing at a velocity of 1 m/s, to have each cross sections on the camera frame appear at an interval of 20 µm, each optical pulse needs to be separated at a time interval of 20 µs. Therefore, a repetition rate of 50 kHz is suitable. Similarly, for a flow velocity of 10 m/s, a repetition rate of 500 kHz is suitable. In both cases, when the field-of-view of the camera is 500 µm, the interval of 20 µm between each cross section results in a maximum of 25 cross sections of the cell.
In addition to the repetition rate, the pulse width must be considered for optimum imaging. A too long pulse width will blur the image in the flow direction due to the motion during the exposure with a single pulse and result in the degradation of resolution; A too short pulse width will reduce the exposure time of a single cross section and result in lower signal-to-noise ratio (SNR) with the same peak power. For a cell flowing at 1 m/s, a pulse width of 200 ns will result in a blur of 200 nm, which is slightly less than the optical resolution of a dry objective lens at visible wavelengths.
2.2 Setup of FLITS
To perform FLITS, we constructed an optical setup as shown in Fig. 2(a). The optical setup mainly consists of two parts: the excitation part and the detection part. In the excitation side, a beam from a 488-nm continuous-wave laser source (Cobolt 06-MLD, HÜBNER Photonics) is shaped into a light sheet using cylindrical lenses (Thorlabs) and then illuminated onto a channel in a quartz flow cell with a channel-cross-section dimensions of 150 × 150 µm (Hamamatsu) using a 10x objective lens (LMPFLN10x, Olympus). To observe the channel from the excitation side, a polarizing beam splitter is placed before the objective lens and the separated light path is guided towards a CMOS camera (DMK 33UX249, Imaging Source). On the detection side, a 20x objective lens with an NA of 0.75 (UPLSAPO20x, Olympus) is directed towards the channel orthogonal to the excitation objective. The collected fluorescence by the 20x objective is imaged with an sCMOS camera (ORCA-Flash4.0 v2, Hamamatsu) equipped with a bandpass filter (ET525/50m, Chroma) in front of it.
Fig. 2. Optical system of FLITS. (a) Optical setup of FLITS. Lens L1, L2, L3 are achromatic lens with focal lengths of 150, 200, and 150 mm, respectively. (b) Positions and angles of channel and light sheet. The channel is tipped in the direction of the detection objective at an angle α to the x-axis. At the same time, the light sheet is rotated in the opposite direction at an angle 0.41 × α to the y-axis. The angle in this figure is enlarged than the actual angle that was used in the experiment for visualization. (c) Image of light sheet taken from the illumination objective side. (d) Profile of light sheet in the z-axis direction. The intensity of a 100-nm fluorescent bead was measured from the illumination objective side while it was translated through the light sheet in the z-axis direction. The FWHM was 1.0 µm. (e) Cross-sectional images of the PSF calculated from the light sheet profile in (d) and the PSF with a wide-field illumination (Fig. S2d). The pixel densities are doubled in the x- and y-axis directions compared to the raw image taken with the sCMOS camera to account for the resolution in the z-direction. The FWHM were 0.88, 0.88, 0.86 µm in the x-, y-, and z-axis directions, respectively. Scale bars: 200 µm in (c); 0.5 µm in (e).
In FLITS, it is necessary to have the light sheet at a slight angle to the channel. This was achieved by slightly tipping the flow cell towards the direction of the detection objective (Fig. 2(b)). However, by changing the angle of the flow cell, the angle of the focal plane of the objective will also change due to the difference in the refractive index of the media between the objective and the flow cell (air, n = 1.0) and inside the channel (water, n = 1.33). Therefore, the angle of the light sheet needs to be changed to match the focal plane. Under the same conditions, when the position of the channel moves d µm towards the objective, the focal position moves 0.41 × d µm away from the objective for an objective with an NA of 0.75 (see Supplemental Document and Fig. S1a). Thus, when the tipping angle of the channel is α (α ≪ 1), the light sheet needs to be rotated 0.41 × α in the opposite angle to the channel (Fig. 2(b) and S1b).
In our setup, the channel was set at an angle of 1.3° and the light sheet was set at an angle of –0.6°. At this angle, when we consider the case when each cross sections are illuminated at 20-µm intervals, the scanning interval along the cell becomes
According to the Nyquist theorem, the axial resolution can be estimated to be 1.34 µm. Together with the width of the light sheet (mentioned below), we can obtain the axial resolution of the system.
After setting up the light-sheet excitation, we evaluated its profile. An image of the light sheet obtained from the excitation side is shown in Fig. 2(c). Because a cylindrical lens was used to expand the light sheet in the vertical (y-axis) direction of the figure, there is a gaussian intensity distribution in this direction. To obtain the thickness of the light sheet, 100-nm fluorescent beads (F8803, Invitrogen) were attached to a glass slide and the intensity of a single bead was obtained while it translated across the light sheet (Fig. 2(d)). The full width at half maximum (FWHM) became 1.0 µm, which corresponds to the thickness of the light sheet.
In addition, we evaluated the resolution of the imaging system. The same 100-nm fluorescent beads were imaged from the detection side with a wide field illumination (Fig. S2a). The FWHM of the spot from a single bead became 0.87 µm (Fig. S2d), which is larger than the resolution of the objective lens with an NA of 0.75. This is because the pixel size is 0.39 µm at the sample and, according to the Nyquist theorem, the resolution would be twice of the pixel size. Although this resolution can be made smaller by changing the pixel size of the camera or the magnification, we use this resolution to keep the field-of-view on the camera large enough to capture 20–30 cell cross sections. Combining this PSF with the profile of the light sheet, we calculated the PSF with the light-sheet excitation as the product of the two, from which we obtained the lateral resolution to be 0.88 µm and the axial resolution to be 0.86 µm (Fig. 2(e)). Because the axial resolution from the Nyquist theorem is larger, the actual axial resolution would be limited to 1.34 µm.
3. Results
3.1 Imaging of hydrogel microspheres
Before imaging objects in flow, we first performed imaging of translating beads on stage with FLITS. Because polystyrene beads refract the light-sheet excitation beam and produces image artifacts [24], we used hydrogel microspheres instead. Here, home-made fluorescent alginate gel microbeads (preparation in Supplemental Document) were trapped in agarose gel inside a 700-µm square capillary with a wall thickness of 140 µm (VitroCom). The capillary was set at an angle of 1.3° and translated in the direction parallel to the capillary at a velocity of 0.5 mm/s. The excitation light was modulated digitally with a repetition rate of 20 Hz and a duty cycle of 1%.
The raw image obtained on the camera is shown in Fig. 3(a). From the above parameters, the z-axis interval of each cross section was 0.80 µm. Using this value, we reconstructed the 3D image of the microbead by stacking each cross section in the z-axis direction. The cross sections in different planes are shown in Fig. 3(c). From these cross sections, the aspect ratio of the image can be seen to be even in all directions. Indeed, the FWHM in the x-, y-, z-axis directions were 14.2, 14.4, 13.1 µm, respectively, which shows the 3D image is nearly spherical. This supports that we are able to obtain the z-axis position correctly using FLITS. The slightly large value in the x- and y-axis directions is likely to be due to the staggering position of each cross sections as can be seen in Fig. 3(c). This occurs because slight variation of velocity during translation. For example, under the conditions for Fig. 3(a) and 3(c), the cross section of the cell will stagger 1 pixel if there is an error in motion of 0.39 µm while traveling 25 µm between each cross section, which is likely to happen if a sample is moved on a motorized stage.
Fig. 3. Images of microspheres obtained with FLITS. (a) Image obtained on the camera for a fluorescent alginate gel microbead on a motorized stage. (b) Image obtained on the camera for a fluorescent PAA bead flowing at 1.07 m/s. (c) yx-, yz-, and xz-cross section views of the 3D reconstructed image for the alginate gel microbead taken in (a). The low intensity region in the z-axis direction are planes that have been photobleached by the light sheet during the adjustment of the imaging system. (d) yx-, yz-, and xz-cross section views of the 3D reconstructed image for the PAA bead taken in (b). Scale bars: 20 µm for (a) and (b); 5 µm for (c) and (d).
Next, we proceeded to imaging hydrogel microspheres in flow. Here, home-made fluorescent poly-acrylamide (PAA) beads (preparation in Supplemental Document) were allowed to flow through the channel of a flow cell with a 150-µm-square cross section (Hamamatsu). The beads were suspended in a sheath buffer (IsoFlow Sheath Fluid, Beckman Coulter) and were hydrodynamically focused with a sheath flow operated at a pressure of 25 kPa and a sample flow rate of 5 µL/min. The excitation light was modulated digitally with a repetition rate of 50 kHz and a duty cycle of 1%.
The raw image obtained on the camera is shown in Fig. 3(b). From this image, the flow velocity of the bead was calculated to be 1.07 m/s from the distance between each cross section, and the z-axis interval of each cross section to be 0.74 µm. With a similar procedure as in the case of beads moving on stage, the 3D image was reconstructed. The cross sections in different planes are shown in Fig. 3(d). Again, we measured the FWHM in the x-, y-, z-axis directions and obtained 11.5, 11.4, and 11.4 µm, respectively, showing a spherical 3D image of the PAA bead. Unlike Fig. 3(c), it can be seen in Fig. 3(d) that there is no staggering of cross sections. This is because the cell is in high-speed flow, and it is difficult for the cell to change its velocity (discussed in Supplemental Document).
3.2 Imaging of cells
We then imaged biological cells in high-speed flow at a velocity of about 1 m/s. Here, we used K562 chronic myelogenous leukemia cells stained with carboxyfluorescein succinimidyl ester (CFSE), which exhibits a green fluorescence upon 488-nm excitation (preparation in Supplemental Document). The cells were suspended in a sheath buffer and passed through a 40-µm-pore cell strainer before flowing. Using the same fluidic system as in the case of PAA beads, the cell suspension was pumped at a flow rate of 10 µL/min while the sheath flow was operated at pressure of 35 kPa. The excitation light was digitally modulated with a repetition rate of 50 kHz and a duty cycle of 1%.
The raw images obtained on the camera are shown in Fig. 4(a). From these images, the flow velocity of the cells was calculated to be 1.34 ± 0.02 m/s for the 5 cells in the figure. From this figure, we analyzed the SNR which we adopt the definition
where µ is the mean intensity of the segmented region (region of the cell) and σ is the standard deviation of the background in the image. The SNR for the 5 cells shown in Fig. 4(a) was calculated to be 200 ± 25. Furthermore, the 3D image was reconstructed as in Fig. 4(b) and Visualization 1.Fig. 4. Images of K562 cells flowing at 1.3 m/s obtained with FLITS. (a) Acquired camera images of K562 cells stained with CFSE. (b) Reconstructed 3D image of the cell in the top panel of (a). Top left, top right, and bottom left are yx-, zx-, and yz-cross sections, respectively, and bottom right is a projected view of a 3D image. (c) Acquired camera images of K562 cells stained with SYTOX Green. (d) Reconstructed 3D image of the cells in (c). Left, center, and right panel groups correspond to the first three images in (c), respectively. Within the panel groups, top left, top right, and bottom left are yz-, zx-, and yz-cross sections, respectively, and bottom right is a projected view of a 3D image. Scale bars: 20 µm for (a) and (c); 5 µm for (b) and (d).
Furthermore, to investigate the ability to image small subcellular structures with the same conditions, we used K562 cells stained with SYTOX Green (Invitrogen), which has similar color fluorescence but bounds to the chromosomes inside the nuclei of cells (preparation in Supplemental Document). From the obtained images in Fig. 4(c), 4(d) and Visualization 2, Visualization 3, and Visualization 4, the structures inside the nuclei caused by the distribution of chromosomes are visible, showing the high spatial resolution of the system. With such resolution, it is clearly distinguishable that, in the last cell in Fig. 4(d), the chromosomes are condensed, and it can be explained to be in the mitotic phase.
Next, we investigated whether FLITS could image cells at a flow velocity even over 10 m/s. First, we used the same K562 cells stained with CFSE, except the sheath flow was operated at a pressure of 180 kPa and the excitation light was digitally modulated with a repetition rate of 500 kHz and a duty cycle of 1%. The raw images obtained on the camera are shown in Fig. 5(a). From these images, the flow velocity of the cells was calculated to be 11.8 ± 0.05 m/s for the 5 cells in the figure. The SNR was calculated to be 21.2 ± 4.9, which is one order of magnitude lower than when the cells flow at 1.3 m/s. Still, the 3D image of the cell can be reconstructed clearly as in Fig. 5(b) and Visualization 5. Then we performed imaging of K562 cells stained with SYTOX Green with the same conditions and observed how the nuclei (or chromosomes) would look. From the images obtained (Fig. 5(c), 5(d) and Visualization 6), the system is able to keep its high spatial resolution even at high flow velocities, with some degradation due to a lower SNR. From these results, it can be concluded that FLITS is able to obtain 3D images of cells flowing at velocities over 10 m/s.
Fig. 5. Images of K562 cells flowing at over 11 m/s obtained with FLITS. (a) Acquired camera images of K562 cells stained with CFSE. (b) Reconstructed 3D image of the cell in the top panel of (a). Top left, top right, and bottom left are yx-, zx-, and yz-cross sections, respectively, and bottom right is a projected view of a 3D image. (c) Acquired camera images of K562 cells stained with SYTOX Green. (d) Reconstructed 3D image of the cells in the top panel of (c). Top left, top right, and bottom left are yx-, zx-, and yz-cross sections, respectively, and bottom right is a projected view of a 3D image. Scale bars: 20 µm for (a) and (c); 5 µm for (b) and (d).
4. Discussion
We have demonstrated that FLITS is able to perform 3D imaging at a flow velocity of 11.8 m/s. Cells with even faster flow velocities can be imaged by using faster modulating lasers. Although at such high flow velocities, the SNR of the image becomes an issue, this can be improved by using pulsed lasers such as Q-switch lasers. The fundamental limit of flow velocity instead arises from the fluorescence lifetime of the fluorophore that is excited. Common fluorophores used to image cells have a fluorescence lifetime of less than 20 ns [25]. If we consider the blur caused by this fluorescence lifetime to be less than the Abbe diffraction limit at a NA of 0.75 and a wavelength of 500 nm, the maximum velocity will be 16.6 m/s. Even if the velocity exceeds this high value, such blur can be eliminated by deconvolution.
Because FLITS obtains cross sections of a single cell in a single frame, the maximum size or the volume of the cell that can be imaged will be limited by the camera frame size and the resolution. For example, in the setup we have used, the camera FOV length is 800 µm and the z-axis cross section interval is 0.74 µm. If we assume the shape of the cell is a sphere with a diameter of R µm, the number of cross sections that can be acquired will be 800/R. Therefore, if the cross sections stacked at 0.74 µm intervals equals the diameter of the cell,
Solving this equation, the maximum size of the cell would be R = 24.3 µm, covering the size of cell types most used in flow cytometry, such as most of immunological cells. For larger cells, either the z-axis interval or the magnification of the objective lens can be changed according to the necessary resolution of the user.
In flow cytometry, the number of cells that can be analyzed in a unit amount of time, or the analysis throughput, becomes important. The throughput of FLITS depends on the speed of the camera, although the velocity of the cell that FLITS can capture is independent of the camera. Because FLITS captures one cell per frame, the maximum possible throughput, if cells arrive in the imaging regions at equal distances between them, equals the frame rate of the camera. When an sCMOS camera is used, the number of active columns on the sensor can be reduced to the size of the cell. If the cell is smaller than 20 µm, the number of columns necessary will be 51 pixels, and a frame rate of over 1,200 frames per second can be achieved (Zyla 5.5 at global shutter mode). This potential throughput is higher than previous reports [23], and with faster cameras [22], even higher throughputs can be achieved.
The advantage of FLITS compared to slow-velocity implementations of 3D imaging flow cytometry is that it becomes compatible with the common commercialized high-throughput cell sorters. These cell sorters work with hydrodynamic flow focusing and require the cells to flow at high velocities of over 1 m/s [11]. To combine 3D imaging with such cell sorting systems, no other methods can perform 3D imaging at such speed, and thus, FLITS becomes necessary. For actual implementation of FLITS in 3D image-based cell sorting, real-time classification of cells based on the raw 2D image may be performed in real-time with GPU and/or FPGA systems as those used for 2D image-based cells sorting [15,16].
In conclusion, we have developed the world’s fastest 3D imaging flow cytometry system based on strobe light-sheet imaging with optofluidic spatial transformation. With the ability to image cells flowing at velocities of over 10 m/s, our method can be implemented on conventional flow cytometry systems including high-throughput cell sorters. We anticipate that our method sheds light onto new possibilities of 3D imaging flow cytometry and high-throughput cell analysis.
Funding
JST, CREST (JPMJCR19H1); JSPS KAKENHI (JP21H00416, JP21H04636); Secom Science and Technology Foundation; UTEC-UTokyo FSI Research Grant Program; Senri Life Science Foundation; Canon Foundation; White Rock Foundation; Uehara Memorial Foundation; Naito Foundation; Noguchi Institute; Tateisi Science and Technology Foundation.
Disclosures
M.U. and S.O. have filed patents related to the FLITS method.
Data availability
Data underlying the results presented in this paper are available in Dataset 1 [26].
Supplemental document
See Supplement 1 for supporting content.
References
1. B.-C. Chen, W. R. Legant, K. Wang, L. Shao, D. E. Milkie, M. W. Davidson, C. Janetopoulos, X. S. Wu, J. A. Hammer, Z. Liu, B. P. English, Y. Mimori-Kiyosue, D. P. Romero, A. T. Ritter, J. Lippincott-Schwartz, L. Fritz-Laylin, R. D. Mullins, D. M. Mitchell, J. N. Bembenek, A.-C. Reymann, R. Böhme, S. W. Grill, J. T. Wang, G. Seydoux, U. S. Tulu, D. P. Kiehart, and E. Betzig, “Lattice light-sheet microscopy: Imaging molecules to embryos at high spatiotemporal resolution,” Science 346(6208), 1257998 (2014). [CrossRef]
2. T. Liu, S. Upadhyayula, D. E. Milkie, V. Singh, K. Wang, I. A. Swinburne, K. R. Mosaliganti, Z. M. Collins, T. W. Hiscock, J. Shea, A. Q. Kohrman, T. N. Medwig, D. Dambournet, R. Forster, B. Cunniff, Y. Ruan, H. Yashiro, S. Scholpp, E. M. Meyerowitz, D. Hockemeyer, D. G. Drubin, B. L. Martin, D. Q. Matus, M. Koyama, S. G. Megason, T. Kirchhausen, and E. Betzig, “Observing the cell in its native state: Imaging subcellular dynamics in multicellular organisms,” Science 360(6386), eaaq1392 (2018). [CrossRef]
3. P. J. Keller, A. D. Schmidt, J. Wittbrodt, and E. H. K. Stelzer, “Reconstruction of zebrafish early embryonic development by scanned light sheet microscopy,” Science 322(5904), 1065–1069 (2008). [CrossRef]
4. E. A. Susaki, K. Tainaka, D. Perrin, F. Kishino, T. Tawara, T. M. Watanabe, C. Yokoyama, H. Onoe, M. Eguchi, S. Yamaguchi, T. Abe, H. Kiyonari, Y. Shimizu, A. Miyawaki, H. Yokota, and H. R. Ueda, “Whole-brain imaging with single-cell resolution using chemical cocktails and computational analysis,” Cell 157(3), 726–739 (2014). [CrossRef]
5. R. Gao, S. M. Asano, S. Upadhyayula, I. Pisarev, D. E. Milkie, T.-L. Liu, V. Singh, A. Graves, G. H. Huynh, Y. Zhao, J. Bogovic, J. Colonell, C. M. Ott, C. Zugates, S. Tappan, A. Rodriguez, K. R. Mosaliganti, S.-H. Sheu, H. A. Pasolli, S. Pang, C. S. Xu, S. G. Megason, H. Hess, J. Lippincott-Schwartz, A. Hantman, G. M. Rubin, T. Kirchhausen, S. Saalfeld, Y. Aso, E. S. Boyden, and E. Betzig, “Cortical column and whole-brain imaging with molecular contrast and nanoscale resolution,” Science 363(6424), eaau8302 (2019). [CrossRef]
6. T. C. George, D. A. Basiji, B. E. Hall, D. H. Lynch, W. E. Ortyn, D. J. Perry, M. J. Seo, C. A. Zimmerman, and P. J. Morrissey, “Distinguishing modes of cell death using the ImageStream® multispectcal imaging flow cytometer,” Cytometry 59A(2), 237–245 (2004). [CrossRef]
7. D. A. Basiji, W. E. Ortyn, L. Liang, V. Venkatachalam, and P. Morrissey, “Cellular Image Analysis and Imaging by Flow Cytometry,” Clinics in Laboratory Medicine 27(3), 653–670 (2007). [CrossRef]
8. N. S. Barteneva and A. I. Vorobjev., Imaging Flow Cytometry: Methods and Protocols (Springer New York, 2016).
9. L. F. Grimwade, K. A. Fuller, and W. N. Erber, “Applications of imaging flow cytometry in the diagnostic assessment of acute leukaemia,” Methods 112, 39–45 (2017). [CrossRef]
10. Y. Han, Y. Gu, A. C. Zhang, and Y.-H. Lo, “Review: imaging technologies for flow cytometry,” Lab Chip 16(24), 4639–4647 (2016). [CrossRef]
11. H. M. Shapiro, Practical Flow Cytometry, 4th ed. (John Wiley & Sons, 2005).
12. K. Goda, A. Ayazi, D. R. Gossett, J. Sadasivam, C. K. Lonappan, E. Sollier, A. M. Fard, S. C. Hur, J. Adam, C. Murray, C. Wang, N. Brackbill, D. Di Carlo, and B. Jalali, “High-throughput single-microparticle imaging flow analyzer,” Proc. Natl. Acad. Sci. 109(29), 11630–11635 (2012). [CrossRef]
13. E. D. Diebold, B. W. Buckley, D. R. Gossett, and B. Jalali, “Digitally synthesized beat frequency multiplexing for sub-millisecond fluorescence microscopy,” Nat. Photonics 7(10), 806–810 (2013). [CrossRef]
14. S. Ota, R. Horisaki, Y. Kawamura, M. Ugawa, I. Sato, K. Hashimoto, R. Kamesawa, K. Setoyama, S. Yamaguchi, K. Fujiu, K. Waki, and H. Noji, “Ghost cytometry,” Science 360(6394), 1246–1251 (2018). [CrossRef]
15. N. Nitta, T. Sugimura, A. Isozaki, H. Mikami, K. Hiraki, S. Sakuma, T. Iino, F. Arai, T. Endo, Y. Fujiwaki, H. Fukuzawa, M. Hase, T. Hayakawa, K. Hiramatsu, Y. Hoshino, M. Inaba, T. Ito, H. Karakawa, Y. Kasai, K. Koizumi, S. Lee, C. Lei, M. Li, T. Maeno, S. Matsusaka, D. Murakami, A. Nakagawa, Y. Oguchi, M. Oikawa, T. Ota, K. Shiba, H. Shintaku, Y. Shirasaki, K. Suga, Y. Suzuki, N. Suzuki, Y. Tanaka, H. Tezuka, C. Toyokawa, Y. Yalikun, M. Yamada, M. Yamagishi, T. Yamano, A. Yasumoto, Y. Yatomi, M. Yazawa, D. Di Carlo, Y. Hosokawa, S. Uemura, Y. Ozeki, and K. Goda, “Intelligent Image-Activated Cell Sorting,” Cell 175(1), 266–276.e13 (2018). [CrossRef]
16. D. Schraivogel, T. M. Kuhn, B. Rauscher, M. Rodríguez-Martínez, M. Paulsen, K. Owsley, A. Middlebrook, C. Tischer, B. Ramasz, D. Ordoñez-Rueda, M. Dees, S. Cuylen-Haering, E. Diebold, and L. M. Steinmetz, “High-speed fluorescence image–enabled cell sorting,” Science 375(6578), 315–320 (2022). [CrossRef]
17. J. Wu, J. Li, and R. K. Y. Chan, “A light sheet based high throughput 3D-imaging flow cytometer for phytoplankton analysis,” Opt. Express 21(12), 14474 (2013). [CrossRef]
18. P. Paiè, F. Bragheri, A. Bassi, and R. Osellame, “Selective plane illumination microscopy on a chip,” Lab Chip 16(9), 1556–1560 (2016). [CrossRef]
19. F. Sala, M. Castriotta, P. Paiè, A. Farina, S. D’Annunzio, A. Zippo, R. Osellame, F. Bragheri, and A. Bassi, “High-throughput 3D imaging of single cells with light-sheet fluorescence microscopy on chip,” Biomed. Opt. Express 11(8), 4397–4407 (2020). [CrossRef]
20. E. Vargas Ordaz, S. Gorelick, H. York, B. Liu, M. L. Halls, S. Arumugam, A. Neild, A. de Marco, and V. Cadarso, “Three-dimensional imaging on a chip using optofluidics light-sheet fluorescence microscopy,” Lab Chip 21(15), 2945–2954 (2021). [CrossRef]
21. M. B. Bouchard, V. Voleti, C. S. Mendes, C. Lacefield, W. B. Grueber, R. S. Mann, R. M. Bruno, and E. M. C. Hillman, “Swept confocally-aligned planar excitation (SCAPE) microscopy for high-speed volumetric imaging of behaving organisms,” Nat. Photonics 9(2), 113–119 (2015). [CrossRef]
22. V. Voleti, K. B. Patel, W. Li, C. Perez Campos, S. Bharadwaj, H. Yu, C. Ford, M. J. Casper, R. W. Yan, W. Liang, C. Wen, K. D. Kimura, K. L. Targoff, and E. M. C. Hillman, “Real-time volumetric microscopy of in vivo dynamics and large-scale samples with SCAPE 2.0,” Nat. Methods 16(10), 1054–1062 (2019). [CrossRef]
23. Y. Han, R. Tang, Y. Gu, A. C. Zhang, W. Cai, V. Castor, S. H. Cho, W. Alaynick, and Y.-H. Lo, “Cameraless high-throughput three-dimensional imaging flow cytometry,” Optica 6(10), 1297 (2019). [CrossRef]
24. F. O. Fahrbach, P. Simon, and A. Rohrbach, “Microscopy with self-reconstructing beams,” Nat. Photonics 4(11), 780–785 (2010). [CrossRef]
25. M. Y. Berezin and S. Achilefu, “Fluorescence lifetime measurements and biological imaging,” Chem. Rev. 110(5), 2641–2684 (2010). [CrossRef]
26. M. Ugawa and S. Ota, “Data and code for “High-speed 3D imaging flow cytometry with optofluidic spatial transformation,” Zenodo, 2022, https://doi.org/10.5281/zenodo.6482119.
