Open Access
Add to BibSonomy
Abstract
Fundamental understanding of large-scale dynamic connectivity within a living organism requires volumetric imaging over a large field of view (FOV) at biologically relevant speed and resolution. However, most microscopy methods make trade-offs between FOV and axial resolution, making it challenging to observe highly dynamic processes at cellular resolution in 3D across mesoscopic scales (e.g., whole zebrafish larva). To overcome this limitation, we have developed mesoscopic oblique plane microscopy (Meso-OPM) with a diffractive light sheet. By augmenting the illumination angle of the light sheet with a transmission grating, we improved the axial resolution approximately sixfold over existing methods and approximately twofold beyond the diffraction limitation of the primary objective lens. We demonstrated a FOV up to ${5.4}\;{\rm mm} \times {3.3}\;{\rm mm}$ with resolution of ${2.5}\;{\unicode{x00B5}{\rm m}} \times {3}\;{\unicode{x00B5}{\rm m}} \times {6}\;{\unicode{x00B5}{\rm m}}$, allowing volumetric imaging of 3D cellular structures with a single scan. Applying Meso-OPM for in vivo imaging of zebrafish larvae, we report here in toto whole-body volumetric recordings of neuronal activity at 2 Hz volume rate and whole-body volumetric recordings of blood flow dynamics at 5 Hz with 3D cellular resolution.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. INTRODUCTION
Modern microscopy, in conjunction with genetically encoded fluorescent reporters, [1–4] has greatly advanced fundamental studies in neuroscience, cardiovascular biology, and developmental biology. As many dynamic biological processes are interconnected across anatomical structures, rapid high-resolution volumetric imaging over a wide field of view (FOV) is critical to interrogating complex molecular and cellular interactions in living organisms [5–7]. For example, neuronal signals often traverse large distances within the brain to coordinate functions [6,7]. However, volumetric recording of neuronal activity patterns with sufficient FOV and resolution is a significant challenge due to the physical limit of optical diffraction. Microscopy techniques using a single primary objective are generally constrained by this limit, such as confocal and multiphoton, dictated by the numerical aperture (NA) of the objective used. This necessitates trade-offs between FOV and spatial resolution. For example, near-centimeter or multi-millimeter FOV can be obtained with subcellular lateral resolution by using a low NA objective lens (e.g., ${\rm NA} = {0.35}$), but the axial resolution is typically above ${\sim}{15}\;{\unicode{x00B5}{\rm m}}$, which is insufficient to resolve cellular features in the $z$ dimension [8–13]. The diffraction limit also extends to microscopy methods that deliver illumination and collect fluorescence via the same refractive optics. For example, a computational miniature mesoscope used a microlens array instead of a regular objective lens to overcome the trade-off between FOV and resolution. Yet, the axial resolution is limited by the total angular coverage of the microlens array leading to axial resolution of tens of micrometers [14].
The diffraction limit also poses an additional challenge in that the axial resolution deteriorates more rapidly than lateral resolution with increasing FOV, since the depth of focus is inversely proportional to the square of the NA. Thus, a much higher NA is needed to achieve axial cellular resolution than cellular lateral resolution. As a reference point, NAs of ${\sim}{0.5 {-} 0.6}$ are needed for two-photon (900 nm excitation wavelength) and single-photon (488 nm excitation wavelength) excitation to achieve cellular axial resolution, respectively [15,16]. Whereas the FOV of a high NA objective lens is fundamentally confounded by the scale-dependent geometric aberrations of its optical elements [17], the increase in NA is at the cost of FOV. Consequently, the FOV is usually less than ${1}\;{\rm mm}\;{\times}\;{1}\;{\rm mm}$ [18–23] when utilizing a high NA objective lens, which is insufficient for imaging dynamic events over large scales. While tremendous efforts have been made to design objectives to achieve large-scale recording with cellular resolution, particularly in the $z$ dimension, the physical limit of diffraction still persists [9,12,15,16], and the complexity of the optics and imaging system also increases dramatically.
Another strategy for overcoming the diffraction limit is to introduce additional optical elements so that resolution is not solely defined by the primary objectives. The prominent example is to use two objectives to decouple excitation and detection as in light sheet microscopy (LSM) [24,25], selective-plane illumination microscopy [26], dual-inverted selective-plane illumination microscopy [27], lattice light-sheet microscopy [28,29], light-sheet theta microscopy [30], and open-top light-sheet microscopes [31]. While the primary objective offers lateral resolution, the excitation lens independently creates a thin optical light sheet and therefore the axial resolution is no longer constrained by the primary objectives.
Because of the diffraction limit, mesoscale volumetric imaging of live samples over multi-millimeter scales with 3D cellular resolution remains a tremendous challenge. In this paper, we introduce a mesoscopic oblique plane microscopy (Meso-OPM) method that employs a novel strategy of a diffractive light sheet to overcome the inherent limitations mentioned above. Meso-OPM belongs to a family generally termed single objective light sheet microscopy (SOLSM) [32,33]. In contrast to conventional LSM, which uses two objective lenses, SOLSM uses only a single primary objective lens but applies an off-axis oblique light sheet excitation at the specimen and a remote focusing system to capture the scanning light sheet. The advantage of SOLSM is the system simplicity in that much of the optics is shared for both excitation and collection, as well as flexible sample mounting as in a conventional upright or inverted microscope setting. Because regular SOLSM uses one primary objective, the resolution limitations described above still apply. Our previous Meso-OPM version achieved ${5}\;{\rm mm}\;{\times}\;{6}\;{\rm mm}$ FOV, but axial resolution is sacrificed to ${\sim}{35}\;{\unicode{x00B5}{\rm m}}$ [13]. Here, the reported Meso-OPM addressed this persistent problem by creating a high-angle illumination light sheet with a transmission grating. This enables the axial resolution to be improved by sixfold, i.e., to cellular level even using a low NA primary objective (e.g., ${\rm NA} = {0.3}$). With 3D cellular resolution, we demonstrate an unprecedented FOV of ${5.4}\;{\rm mm} \times {3.3}\;{\rm mm}\times \;{0.33}\;{\rm mm}$ in an acute brain slice preparation at a resolution of ${2.5}\;{\unicode{x00B5}{\rm m}} \times {3}\;{\unicode{x00B5}{\rm m}} \times {6}\;{\unicode{x00B5}{\rm m}}$, to resolve individual neurons. Using Meso-OPM to image living larval zebrafish, we also demonstrated whole-brain cellular resolution imaging of calcium dynamics at 2 Hz and whole-body blood flow imaging at 5 Hz.
2. MATERIALS AND METHODS
A. Working Principle of Meso-OPM
The limitation of previous Meso-OPM methods is the inherent trade-off between FOV and resolution, particularly in the depth direction [10,13]. For a low NA objective (e.g., ${\rm NA} = {0.3}$), the axial resolution is on the order of tens of micrometers. To circumvent this trade-off, we introduced a diffractive light sheet by using a transmission grating [Fig. 1(a)]. A large angle diffractive light sheet can be created in the first order while maintaining the full collection angle for emission collection in the zeroth order. The diffraction by the grating is independent of the primary objective lens, and thus not constrained by the diffraction trade-off discussed above. The concept for the proposed Meso-OPM system is shown in Fig. 1(b). The light sheet angle mainly relies on the diffraction angle such that the axial resolution is improved sixfold [10,13] compared to our previous mesoscopic OPM without grating [Fig. 1(c)] and twofold of the diffraction-limited axial resolution (Figs. S1 and S2), thus achieving 3D cellular resolution. We compared the volumetric FOV (VFOV) and the focal volume (FV) by the proposed Meso-OPM with an array of SOLSM techniques [10,13,19,32–40] [Fig. 1(d)], where FV is estimated by the product of the resolution in $X$, $Y$, and $Z$. The volumetric throughput can be assessed by VFOV/FV, where our system achieved ${\sim}{1.3} \times {{10}^8}$, more than three times the state-of-the-art value at ${3.5} \times {{10}^7}$ [10]. The detailed comparison of FOV and resolution is summarized in Table 1. It can be found that our method maintains multi-millimeter FOV as well as micrometer-level axial resolution. Consequently, Meso-OPM can afford capturing 4D biological dynamics at cellular resolution across FOV dimensions that have been challenging to achieve previously.
Fig. 1. Concept of volumetric imaging with a diffractive light sheet. (a) High-angle light sheet illumination in first order and full NA fluorescence collection in zeroth order with the help of transmission grating. (b) Concept of utilizing diffractive light sheet in Meso-OPM to achieve large FOV and 3D cellular resolution. (c) Comparison of spatial resolution with existing mesoscopic OPM. (d) Comparison of the volumetric FOV and focal volume (FV) between different SOLSMs.
Table 1. Comparison of the Lateral FOV, Axial Resolution, VFOV, and Focal Volume (FV) between Different SOLSMsa
B. System Description
The layout of the optical design for Meso-OPM is shown in Fig. 2(a) (see supplementary Fig. S3 for actual setup photograph). The key component that enables 3D cellular resolution over large FOV in Meso-OPM is the transmission grating (TG) shown in Fig. 2(b). The excitation light sheet (blue) in Fig. 2(b) can be diffracted into different orders by the flowing equation:
where ${\theta _m}$ is the diffraction angle for the $m$th order, $m$ represents the diffracted order, $\lambda$ is the wavelength of the incident beam, ${\theta _i}$ is the incident angle, and $d$ is the line density of the TG. We used the first-order diffraction for excitation, and collected the zeroth-order fluorescence emission. Fluorescence in nonzero orders is either rejected by the aperture of the objective lens (OL1) or widely separated in the image space [Fig. 2(c) and supplementary Fig. S4]. With a larger illumination angle by the grating diffraction (see supplementary Fig. S5), the axial resolution is improved beyond the limitation set by the NA of the primary objective lens [see supplementary Figs. S6(d) and S6(e)]. Having described the mechanism of creating and imaging with the high-angle diffractive light sheet, we set out for detailed explanations of the whole system below.Fig. 2. Schematic of the experiment setup. (a) Layout of the whole optical design. TG, transmission grating; GW, glass window; AF, aluminum foil; OL, objective lens; L, lens; M, mirror; GM, galvanometer mirror; IIP, intermediate image plane; F, filter; TS, translation stage; PL, Powell lens; FM, flip mirror; LS, light source. (b) Zoom-in view of the primary objective lens exhibiting the generation of high-angle excitation light sheet. The primary objective lens consists of the objective lens (OL1), transmission grating (TG), glass window (GW), and motorized aluminum foil (AF). (c) Simplified layout of the optics from the light sheet to the IIP showing the imaging of emission fluorescence in different diffraction orders.
The light source unit consists of two lasers (LS1: Coherent, 488 nm, 50 mW; LS2: Coherent, 561 nm, 100 mW), a Powell lens (PL: Edmund optics, 30º fan angle), a telescope (L5: Thorlabs, two AC254-250-A, ${f_{\rm L5}} = {125}\;{\rm mm}$; L6: two AC254-050-A, ${f_{\rm L6}} = {25}\;{\rm mm}$), two plane mirrors (M2 & M4: Thorlabs, PF10-03-P01), and one flip mirror (FM3: Thorlabs, FP90, and PF10-03-P01). Excitation light can be switched between LS1 and LS2 by FM3. PL has a fun angle of 30º and is used to create light sheet excitation. The light sheet was introduced into the microscope by a dichroic mirror (DM, Chroma, ZT488/561rpc-UF1). The light source unit is mounted on a translational stage to adjust the laser offset from the optical axis for oblique illumination.
The laser is then directed by two relay telescopes (L4, L3, L2: Thorlabs, two AC508-200-A, ${f_{\rm L4}} = {f_{\rm L3}} = {f_{\rm L2}} = {100}\;{\rm mm}$; L1: Thorlabs, two AC508-100-A, ${f_{\rm L1}} = {50}\;{\rm mm}$) and a scanning galvanometer mirror (Nutfield: QS-12 OPD, 20 mm aperture) to the back pupil of the primary objective (OL1, Olympus, UplanFL ${10} {\times} { /0.3}$, ${f_{\rm OL1}} = {18}\;{\rm mm}$). The offset of the beam at the back pupil of OL1 is ${\sim}{3.5}\;{\rm mm}$. To reduce the variation of the excitation light sheet angle, the rotation axis of the galvanometer mirror should be conjugated to the intersection of the optical axis and the back focal plane of the primary objective lens. According to the calculation in [41], the variance of the light sheet at the sample is ${\sim}{0.25}^\circ$ when the scanning angle $( {\lt} {5}^\circ)$ covers the full FOV. As a result, the light sheet has a near constant incident angle of ${\sim} {11}^\circ$ when projected on the transmission grating (TG: Thorlabs, line density of 0.83 µm/line). The angle of the first-order diffraction in the air can be calculated by Eq. (1) to be ${\sim} {51}^\circ$ and ${\sim }{60}^\circ$ for 488 nm and 561 nm, respectively. The oblique angle ${\theta _r}$ in water is then ${\sim }{36}^\circ$ and ${\sim }{41}^\circ$ correspondingly by refraction [see supplementary Fig. S5(a)]. The diffraction efficiencies in the first order for 488 nm and 561 nm excitation are measured to be ${\sim}{26}\%$ and ${\sim}{20}\%$, while the diffraction efficiencies in the zeroth order for the green and red emission are both ${\sim}{35}\%$.
Beyond the transmission grating, two coverslips were used to protect grating surface and for water immersion [Fig. 2(b)]. The working distance from the OL1 assembly is ${\sim}{1 {-} 2}\;{\rm mm}$. The distance between zeroth-, first-, and ${-}{1}$st-order excitation is ${\sim}{1}\;{\rm mm}$. As long as the sample size is ${\lt}{1}\;{\rm mm}$, no interference from zeroth- and ${-}{1}$st-order excitation is present. For imaging a large specimen, such as a brain slice, the zeroth and ${-}{1}$st orders of the diffraction were blocked by a moving aluminum foil Visualization 5 (AF) sandwiched between two coverslips [Fig. 2(b) and supplementary Figs. S3(b) and S3(c)] to avoid undesired fluorescence excitation.
As for excitation wavelength of 488 nm, the beam waist (${\omega _0}$) and the Rayleigh (${Z_R}$) range in the immersion water are approximately 5 µm and 214 µm, respectively. As for 561 nm, the beam waist and Rayleigh range are slightly bigger, which are 6 µm and 267 µm, respectively. [See the calculation with Eqs. (S2) and (S3) in Supplement 1). The imaging depth range is then estimated by ${2}{Z_{R\:}} \cos{\theta _r}$, where ${\theta _r}$ is the oblique angle of the light sheet in water [i.e., 36° and 41° for 488 and 561 nm; see supplementary Fig. S5(a)], to be 346 and 403 µm.
The zeroth-order fluorescence emission through the grating is collected by the system. The light sheet image is directed by the relay lenses L1 and L2 and descanned by GM. A stationary intermediate image plane (IIP) can be formed after OL2 (Olympus, UPLSAPO ${20} {\times} { /}0.75$, ${f_{\rm OL2}} = {9}\;{\rm mm}$). The additional glass substrate of the transmission grating, as well as the immersion water, may increase spherical aberration. To mitigate spherical aberration, the magnification from L1 to L4 was designed to be 2 (${f_{\rm L2}} \times {f_{\rm L4}}/{f_{\rm L1}}/{f_{\rm L3}} = {2}$) so that the back aperture of the OL2 is overfilled to reject high-angle rays at the cost of light loss [see supplementary Eq. (S6) and Fig. S7]. According to the simulation in Fig. S8, the optical system from OL1 to OL2 is less sensitive to glass and immersion media induced aberration. The optical performance can be significantly improved by better optical design and customized transmission grating, which will be discussed in the end.
The lateral magnification ${M_L}$ from the specimen to IIP can be calculated according to the focal length of lenses as ${M_L} = {f_{\rm OL2}} \,\times\def\LDeqbreak{} {f_{\rm L1}} \times {f_{\rm L3}}/{f_{\rm OL1}}/{f_{\rm L2}}/{f_{\rm L4}} = {0.25}$. The axial magnification ${M_a}$ is ${0.{25}^2}/{1.33}$ (1.33 is the refractive index of the immersion media). Given the magnification (${M_L} = {0.25}$) and the angle of the light sheet (${\theta _r} = {36}^\circ$ for 488 nm, 41° for 561 nm), the angle (${\theta _{{\rm IIP}}}$) of the light sheet image in IIP can be calculated as follows [13]:
Fig. 3. High-definition whole-body vasculature imaging of live zebrafish larva obtained in a single FOV at frame rate of 500 Hz (see Table S1 for imaging parameters). (a) Color-coded maximum intensity projection (MIP) in the $XY$ plane over 425 µm along the $Z$ direction. The 0 reference position of the MIP is marked by a red arrow in panel (b). (b) Color-coded MIP in the $YZ$ plane over 640 µm along the $X$ direction indicated by the yellow dashed box in panel (a). The 0 reference position of the MIP is marked by a red in panel (c), (c) Color-coded MIP in $XZ$ plane over 500 µm along the $Y$ direction. The 0 reference position of the MIP is marked by a red arrow in panel (a). (d) $YZ$ cross sections of four pairs of ISVs indicted by corresponding double dashed lines in panel (c). (e)–(j) Enface $z$ projections of the head region marked by yellow dashed box in panel (a). Paired vessel structures such as MCeV (middle cerebral vein), PHBC (primordial hindbrain channel), CCtA (cerebellar central artery), PMCtA (posterior mesencephalic central artery), AMCtA (anterior mesencephalic central artery), and PCS (posterior communicating segment) can be seen within different projections. Scale bar, 100 µm. (BCA, basal communicating artery; BA, basilar artery).
As for excitation wavelengths of 488 nm and 561 nm, ${\theta _{{\rm IIP}}}$ is calculated to be ${\sim }71^{\circ}$ and 74º in air, respectively. Finally, a remote focusing system, consisting of an objective lens (OL3: Olympus, ${\rm UPLFLN}{40} {\times} { /}0.75$, ${f_{\rm OL3\:}} = {4.5}\;{\rm mm}$), filter (F1: Chroma, MF525-39; Thorlabs, AT575lp), camera lens (L7: Navitar, MVL75M1, ${f_{\rm L7}} = {75}\;{\rm mm}$), and camera (Andor, Zyla 4.2), is correspondingly rotated by 19º or 16º for collection. Because of the augmented light sheet angle by grating, the IIP is almost perpendicular to the optical axis [zoom-in of the IIP in Fig. 2(a)]. Therefore, ${\gt}{80}\%$ of the light from OL2 to OL3 can be collected, resulting in an effective NA (lateral) of ${\sim}{0.2}$ (see Fig. S7).
C. Theoretical Resolution
After the description of the setup, we now give the theoretical expression of the system resolution. As the angle of the excitation light sheet is independent of the NA of the primary objective lens (OL1), the calculation of the system point spread function (PSF) can be written as follows:
where ${{\rm PSF}_{{\rm Ill}}}$ is the illumination PSF, and ${{\rm PSF}_{{\rm Det}}}$ is the detection PSF. The calculation is the same as that for the LSM except for the nonorthogonality between the illumination and detection PSF. A numerical simulation based on our previous publications [42,43] is used to calculate the theoretical resolution according to Eq. (3) [see supplementary Fig. S6(a)]. We first established the 3D coherent amplitude transfer function, and then simulated PSFs for both illumination and emission. The calculated theoretical resolutions are ${\sim}{1.4}\;\unicode{x00B5}{\rm m}\;({X}) \times {1.6}\;\unicode{x00B5}{\rm m}(Y)\times 6\,\,\unicode{x00B5}{\rm m}(Z)$ for 488 nm excitation and ${\sim}{1.5}\;\unicode{x00B5}{\rm m}\;({ X}) \times {1.8}\;\unicode{x00B5}{\rm m}\;({Y}) \times {6}\;{\unicode{x00B5}{\rm m}}\;({ Z})$ for 561 nm excitation.D. Data Processing
As a result of the oblique illumination light sheet, the acquired image volume is a stack of oblique 2D images. Therefore, affine transformation consisting of both shearing and scaling is applied to the volume data to reconstruct the actual geometry of the sample [10,34,42].
As for Fig. 5(b) and Visualization 4, the lower part of the fish data (mainly the digestive system) is removed to highlight the neuronal features in the brain and the spinal cord. To generate the 2D angiogram shown in Fig. 6(d)–6(g), every two adjacent volume data sets in the whole time series are subtracted to first generate a differential time series. Then, a 3D angiogram of the entire fish can be created by applying a maximum intensity projection (MIP) along the time axis of the differential time series. Lastly, the projection [Fig. 6(d)–6(g)] was obtained by taking the MIP along each dimension ($X$, $Y$, and $Z$) of the 3D angiogram.
E. Biological Sample Preparation
All animal-related procedures were in accordance with the Institutional Animal Care and Use Committee at Johns Hopkins University and conformed to the guidelines on the Use of Animals from the National Institutes of Health (NIH).
Red and green fluorescent microspheres (Polysciences: YG Microspheres, 1.00 µm; Red Dyed Microsphere, 1.00 µm) were used in the experiments for resolution characterization. They were diluted and immobilized in 1% agarose and molded in two separate Petri dishes. The Petri dish was covered with a cover glass (thickness, 0.1 mm; refractive index, ${\sim}{1.52}$) to flatten the gel surface. Water immersion is applied between the protective glass window of the objective lens and the cover glass before imaging.
Transgenic zebrafish larvae expressing green reef coral fluorescent protein in endothelial cells (Tg(VEGFR2:GRCFP)ZNL, Fig. 3), green fluorescent protein in macrophage cells (Tg(spi1b:GAL4,UAS:EGFP), Fig. 6), and genetically encoded calcium indicators (Tg(elavl3:jGCaMP7s), Fig. 5) were used in the in vivo experiments. Zebrafish at 4–5 dpf with length of 3–4 mm were cultivated at ${\sim}{28}^{ \circ}\rm C$ following standard procedures (light cycle conditions of 14 h light and 10 h dark). To mount samples, zebrafish larvae were first anesthetized with tricaine (MS-222) for ${\sim}{3}\;{\rm min}$ and then added to heated 1.5% low-melting agarose in a Petri dish to allow for orientation at room temperature before providing overlayed E3 media following agarose hardening. The refractive index of the water is 1.33. The objective was immersed into the Petri dish to image the fish directly in an upright way. After the imaging, all of the zebrafish were released into freshwater for recovery.
Fig. 4. High-definition 3D imaging of an acute brain slice obtained in a single FOV at 3D cellular resolution across a wide FOV (see Table S1 for imaging parameters). (a) Color-coded MIPs of the enface view over a distance of 250 µm in the $Z$ dimension, i.e., the whole brain slice volume. (b) Color-coded MIPs of the $XZ$ plane over a distance of 450 µm in the $Y$ dimension. The 0 reference positions of the projections are indicated by blue dashed lines in panel (a). (c),(e),(g) Zoomed images of the enface views of the areas indicated by dashed boxes in panel (a). (d),(f),(h) Color-coded MIPs of the $YZ$ plane of the volume indicated by double arrow lines in panels (c), (e), and (g). The scale bar for panel (a) is 200 µm; the scale bars for the other panels are all 100 µm.
For the in vitro slice imaging, progeny from homozygous Calb2-IRES-cre mice (JAX no. #010774) crossed with homozygous tdTomato reporter mice (JAX no. #007909) was used. To prepare the acute brain slice, the whole brain of a four-week-old mouse was dissected out in artificial cerebrospinal fluid ACSF containing (in mM) 130 NaCl, 3 KCl, 1.25 NaH2PO4, 20 NaHCO3, 10 glucose, 1.3 MgSO4.7H2O, and 2.5 CaCl2.2H2O (pH 7.35–7.4, equilibrated with 95% O2–5% CO2). A 400-µm tissue block containing both the thalamus and cortex region [44] was sliced in icy-cold oxygenated ACSF, using a microtome (Leica VT1200), and transferred into oxygenated ACSF warmed at physiological temperature (35–37°C) for an hour. After recovery, the acute brain slice was transferred to a perfusion chamber cycled with oxygenated ACSF at room temperature for imaging. The refractive index of the ACSF is 1.33. The objective lens was immersed into the perfusion chamber during the imaging.
3. RESULTS
For each demonstration shown below, the imaging parameters are summarized in Supplement 1 (Table S1), while the sample preparation is provided in Section 2.
A. FOV and Resolution Characterization
As the angle of the diffractive light sheet is also a function of the excitation wavelength, we performed imaging experiments on two types of fluorescent microspheres to calibrate the FOV and resolution under blue and green light excitation. The lateral scale was calibrated by imaging the resolution target. To calibrate the axial scale, the sample was moved in the $Z$ dimension by a manual positioning stage during imaging. The FOV under different excitation wavelengths is the same, which is ${3.3}\;{\rm mm}\;({X}) \times {5.4}\;{\rm mm}\;({Y}) \times {0.33}\;{\rm mm}\;({Z})$ (Supplement 1, Figs. S1 and S2) obtained at 1 mm working distance. The resolution was quantified by calculating the full width at half maximum (FWHM) of the line intensity along the $X$, $Y$, and $Z$ directions of different beads. The resolutions along the $X$, $Y$, and $Z$ directions for blue light (488 nm laser) excitation are ${2.5}\;{\pm}\;{1.1}\;{\unicode{x00B5}{\rm m}}$, ${3}\;{\pm}\;{1.4}\;{\unicode{x00B5}{\rm m}}$, and ${6}\;{\pm}\;{1.8}\;{\unicode{x00B5}{\rm m}}$, respectively, while those for green light (561 nm laser) excitation are ${2.9}\;{\pm}\;{1}\;{\unicode{x00B5}{\rm m}}$, ${3.5}\;{\pm}\;{1.5}\;{\unicode{x00B5}{\rm m}}$, and ${6}\;{\pm}\;{1.9}\;{\unicode{x00B5}{\rm m}}$, respectively (Supplement 1, Figs. S1 and S2). By dividing the accessible FOV of ${3.3}\;{\rm mm}\; \times {5.4}\;{\rm mm} \times \;{0.33}\;{\rm mm}$ by the volumetric resolution of ${2.5}\;{\unicode{x00B5}{\rm m}}\; \times {3}\;{\unicode{x00B5}{\rm m}} \times \;{6}\;{\unicode{x00B5}{\rm m}}$, the information throughput of our Meso-OPM can be estimated to be ${\sim}{1.3} \times {{10}^8}$ resolvable image points across the imaged volume. By extension, this suggests the ability to monitor 4D signal dynamics of ${\sim}{1.3} \times {{10}^7}$ cells.
B. High-Resolution Structure Imaging in Whole Larval Zebrafish and Uncleared Acute Mouse Brain Slice
1. Whole Larval Zebrafish Imaging With Partial FOV
After characterizing the FOV and resolution, we carried out static structural imaging of large living samples to test the system performance. We first imaged transgenic zebrafish larvae that express a green fluorescent protein reporter in vasculature endothelial cells at four days post fertilization (dpf). The approximate dimensions of a zebrafish larvae at this age are ${\sim}{3 {-} 4}\;{\rm mm}$ long, ${\sim}{0.6}\;{\rm mm}$ wide, and ${\sim}{0.5}\;{\rm mm}$ thick [45]. The laser power on the sample was ${\sim}\;{3.4}\;{\rm mW}$. The acquired volume size has dimensions of ${3.3}\;{\rm mm}\;({X}) \times {0.65}\;{\rm mm}\;({ Y}) \times {0.55}\;{\rm mm}\;({Z})$ with pixel density of ${2048} \times {300} \times {275}$ pixels at a frame rate of ${\sim}{500}\;{\rm Hz}$. The image quality is best within the calibrated FOV of ${\sim}{0.33}\;{\rm mm}$ in the $Z$ dimension. By achieving 3D cellular resolution over a wider and deeper FOV, we can provide a high-definition 3D rendering of the whole larva (see Visualization 1).
Color-coded projections of volumetric data demonstrate the ability of Meso-OPM to resolve individual blood vessels throughout larval zebrafish in 3D [Figs. 3(a)–3(c)]. The branches of the mesencephalic vein (MsV) in the dorsal head region and the intersegmental veins (ISV) in the ventral trunk can be observed in both the lateral and axial projections [Figs. 3(a)–(c)]. Thanks to the large FOV, the whole dorsal longitudinal anastomotic vessel (DLAV), which has a length of ${\sim}\;{3}\;{\rm mm}$, can be fully visualized in the lateral [Fig. 3(a)] and axial views [Fig. 3(c)]. The DLAV is connected to either the dorsal aorta (DA) or posterior (caudal) cardinal vein (PCV) by ${\sim}{29}$ pairs of ISVs over a distance of ${\sim}{3}\;{\rm mm}$ [Fig. 3(c)]. The ability to discriminate all ISVs demonstrates that a depth penetration of ${\sim}{250}\;{\unicode{x00B5}{\rm m}}$ can be maintained throughout the FOV [Fig. 3(c)]. The intensity profile of the small feature along the depth direction suggests axial resolution better than 6.5 µm [white dashed box shown in Fig. 3(c)], which is sufficient to resolve individual vessels in the depth direction. The double-edge structure of the ISV vessels can be revealed by the intensity profile of a single ISV [blue dashed box shown in Fig. 3(c)]. Transverse views of the trunk at four rostral–caudal positions clearly reveal the round shape of the neural tube (NT) and notochord (NC) [Fig. 3(d)], which typically have diameters of ${\sim}{25}\;{\unicode{x00B5}{\rm m}}$ and 50 µm, respectively [46]. Enface $z$ projections of the head region at six different depths demonstrate the ability to resolve individual vessels throughout the imaging volume [Figs. 3(e)–3(j)]. The discrimination of different vessel patterns, such as the paired brain vascular structures, indicates the axial resolution and penetration can be well maintained in the peripheral area of the FOV.
2. Uncleared Acute Mouse Brain Slice Imaging with Full FOV
We next imaged an acute mouse brain slice with calretinin neurons expressing tdTomato under the full imaging volume of ${3.3}\;{\rm mm}\;({ X}) \times {5.4}\;{\rm mm}\;({Y}) \times {0.33}\;{\rm mm}\;({ Z})$ with a pixel density of ${2048} \times {2000} \times {160}$ pixels at frame rate of 100 Hz. The laser power on the sample was ${\sim}{2}\;{\rm mW}$. To eliminate the background by the zeroth- and $-1$st-order diffraction by the grating, a synchronized moving aluminum foil is used as a beam blocker allowing only first-order diffraction excitation during the acquisition (see Fig. S3d in Supplement 1). Figure 4(a) shows a single scan color-coded mesoscopic FOV covering major anatomical structures such as the cortex, hippocampus, and thalamus. The uniform somatic structures indicate that the cellular resolution is well maintained over the full FOV. Neuronal somas are clearly visualized from cortical layer I to IV in all three projections [Figs. 4(a)–4(d)], confirming 3D cellular resolution. The large FOV allows us to simultaneously capture neurons in the hippocampus [Figs. 4(e) and 4(f)] and the thalamus [Figs. 4(g) and 4(h)]. Close examination demonstrates the capability to image apical dendrites of pyramidal neurons in layer VI [Fig. 4(i)]. A high-definition 3D rendering of the acquired volume can be found in Visualization 2.
Fig. 5. Whole-body neuronal activity recording in zebrafish larva with 3D cellular resolution (see Table S1 for imaging parameters). (a) Color-coded MIP in $XY$ plane over 250 µm along the $Z$ dimension. The reference positions of the MIP are marked by arrows in panel (b). (b) Color-coded MIP in $XZ$ plane over 210 µm in the $Y$ dimension. The reference positions of the MIP are marked by arrows in panel (a). (c)–(f) Enface projections at different $z$ depths across the whole brain. (g)–(j) Zoomed views of the boxed areas in panels (c)–(f). (k) Representative neurons selected for evaluating correlated activity patterns are marked with circles; the color represents relative imaging depth from 0 to 120 µm. Fish data shown in (a)–(k) were acquired at 250 Hz per plane and scanned 125 planes at ${\sim}{4}\;{\unicode{x00B5}{\rm m}}$ spacing, thus a 2 Hz whole-volume resolution. (l) Calcium traces of the selected neurons [see panel (l)] over an 85 s recording period. Scale bar, 100 µm.
C. Whole-Body Recording of Neuronal Activity in Zebrafish Larvae
After characterizing the FOV and resolution in static images of living samples, we proceeded to image neuronal signaling dynamics across an entire zebrafish larva expressing a cytosolic genetically encoded calcium indicator (jGCaMP7s) in neurons throughout the central nervous system [47,48]. The acquired volume of the whole 4-dpf zebrafish has dimensions of ${\sim}\;{3.3}\;{\rm mm} \times {0.5}\;{\rm mm} \times {0.4}\;{\rm mm}$. The laser power on the sample was ${\sim}\;{3.4}\;{\rm mW}$ with which no significant photon bleaching was observed during the imaging process. Color-coded MIPs of the whole zebrafish larvae are shown in Figs. 5(a)–5(c). The entire volume data has a pixel dimension of ${2048} \times {125} \times {200}$ pixels (each volume contains 125 planes) at 2 Hz volume rate. Thanks to the large FOV, individual neuronal features can be identified throughout the entire nerve system from the brain to the spinal cord, as shown in the enface projection [Fig. 5(a)], as well as the side projection [Fig. 5(b)]. MIPs of different brain regions at varying depths were generated [Figs. 5(c)–5(f)]. Distinct features can be clearly delineated, such as the boundary of the hemispheres indicated by the yellow arrows in Figs. 5(c)–5(f) and the boundary of the optic tectum in Fig. 5(g). For instance, tectal neurons in the midbrain can be clearly identified [Figs. 5(c), 5(g), 5(d), and 5(h)], as well as neurons in the cerebellum [Figs. 5(e) and 5(i)]. Pallium and habenula neurons that are not present in Figs. 5(c)–5(e) can be seen in a deeper layer MIP [Figs. 5(f) and 5(j)], which agrees with the anatomy structure of the zebrafish brain [49,50]. Zoomed images of the boxed areas in Figs. 5(g)–5(j) further confirm that individual neurons can be resolved at different tissue depths. A comparison of the image acquired with the Meso-OPM and the confocal microscope is provided in Supplement 1, Fig. S13, showing the similar anatomical organization of neurons.
Fig. 6. Imaging blood flow over the whole larval zebrafish at high spatiotemporal resolution (see Table S1 for imaging parameters). (a) Single volume captured at volume rate of 1 Hz to highlight the structure. (b) Color-coded MIP of cell movements in the $XY$ plane. The $XY$ view is generated by spatial MIP along the $Z$ direction over the position indicated by blue arrows in panel (c). (c) Color-coded MIP of cell movements in the $XZ$ plane. The $XZ$ view is generated by spatial MIP along the $Y$ direction over the position indicated by blue arrows in panel (b). (d) Color-coded angiogram in $XY$ plane. (e)–(g) Tracking of blood cells in 3D over a time window of 4 s in a single ISV indicated in panel (e). Fish data shown in (b)–(g) were acquired at 625 Hz per plane and scanning 125 planes at ${\sim}{4}\;{\unicode{x00B5}{\rm m}}$ spacing, thus a 5 Hz whole-volume resolution. (h) Measurement of velocities of blood cells in different types of vessels. The heart rate was estimated to be ${\sim}{140}\;{\rm Hz}$ by Fourier analysis. Scale bar, 100 µm.
Spontaneous calcium dynamics over the entire zebrafish nervous system were captured at a volume rate of 2 Hz for 85 s. Sixty neurons were manually selected at different depths throughout the brain and spinal cord [Fig. 5(k)] and fluorescence intensity traces plotted [Fig. 5(l)]. These data show that at two time points many of the selected neurons appeared to fire synchronously. This correlated firing pattern can also be observed in Visualization 3. As the FOV covers the whole larva, Meso-OPM offers a unique opportunity to study how neuronal activities are correlated over large volumes encompassing the entire brain and spinal cord, a level of inquiry that has been inaccessible to date. A 4D rendering of the time series can also be found in Visualization 3.
D. Imaging Blood Flow over the Whole Larval Zebrafish at High Spatiotemporal Resolution
By achieving large-field 3D cellular resolution across entire sample volumes without moving the sample or objective lens, our novel Meso-OPM method is well-suited for imaging cellular dynamics across whole organisms or whole tissues at fast speeds. As an additional demonstration of this, we performed volumetric imaging of blood flow using a transgenic zebrafish expressing enhanced green fluorescent protein (EGFP) in myeloid cells. The laser power on the sample is ${\sim}\;{3.4}\;{\rm mW}$. The structure of the entire larval fish was first confirmed by the single volume shown in Fig. 6(a). Next, images were acquired across a whole 4-dpf zebrafish larva (a ${\sim}{3.3}\;{\rm mm} \times {0.5}\;{\rm mm} \times {0.4}\;{\rm mm}$ volume) with pixel dimensions of ${2048} \times {125} \times {200}$ pixels at a volume rate of 5 Hz (see Visualization 4). Results of timelapse imaging over a 9 s time window show individual cell dynamics within major vessels of the circulatory system, such as the CA, caudal vein (CV), DLAV, ISV, posterior (caudal) cardinal vein (PCV), and DA [Figs. 6(b) and 6(c)]. Cell movements are indicated by color-coded temporal traces [Figs. 6(b) and 6(c)]. The discrimination of individual blood cell movements in 3D across the whole volume, from heart to tail, further demonstrates the high temporal and spatial resolution possible with Meso-OPM (see 4D rendering in Visualization 4). To quantify blood flow, motion contrast angiography (see Section 2) was performed on the whole time series to highlight the motion and suppress the static signals as shown in Fig. 6(d).
To exemplify the power of the high spatiotemporal resolution achieved, the 4D trajectory of a single cell traveling within an ISV from the DLAV to the CV over a time window of 4 s is provided in Figs. 6(e)–6(g). With the capability of tracking individual cells, we are able to quantify the speed of blood flow in different types of vessels such as the DA, PCV, and ISV [Fig. 6(h)]. The measured results agree with previous publications [51] revealing different velocities in each vessel type. For example, the velocity of the blood cell in the DA is faster than in the PCV or ISV as expected. Similarly, the pulsatile movement of the blood cells in the DA and PCV is more apparent than in ISV. Owing to the large FOV, blood cells in the DA were tracked over a distance of ${\sim} 2\,\,\rm nm$ from the heart to the tail. An interesting observation is that the velocity in the DA gradually slows down as the blood cells are traveling away from the heart.
Fig. 7. Change of FOV and axial resolution with different imaging parameters. (a) Change of the axial resolution and the imaging depth range with that of the excitation angle under 488 nm excitation. (b),(c) Change of FOV in the $Y$ direction under different working distances (WD).
4. DISCUSSION
Our diffractive mesoscopic OPM leverages the large FOV of the off-the-shelf low magnification objective lens (e.g., 0.3 NA) while overcoming the limitation of insufficient axial resolution. The use of a diffractive light sheet in Meso-OPM allows 3D cellular resolution over a ${3.3}\;{\rm mm} \times {5.4}\;{\rm mm}$ FOV without any mechanical translation. We demonstrated the performance by structural imaging of entire zebrafish larvae and a large acute brain slice. With large FOV and 3D cellular resolution, and biologically relevant speed, we were able to record calcium dynamics over the entire zebrafish nerve system at volume rate of 2 Hz, which was previously unattainable due to insufficient axial resolution [10,13] or limited FOV covering only the brain [18,25]. We also demonstrated single blood cell 4D tracking at volume rate of 5 Hz over the whole larval zebrafish. Imaging and quantifying dynamic cellular and signaling events between multiple tissues/regions, such as the recording of spontaneous neuronal activity between the brain and spinal cord, and blood cell circulation between multiple vessel types (e.g., from the DA to CA), has been an unmet challenge that Meso-OPM can overcome.
Due to the geometry of the diffractive light sheet, the current system makes trade-offs among the working distance, FOV, and axial resolution. With the theoretical framework and the numerical simulation established in Supplement 1, we calculated the axial resolution under different excitation angles by changing the grating line density. As shown in Fig. 7(a), a better axial resolution can be achieved by a larger excitation angle at the expense of a shorter imaging depth range. Meanwhile, a large working distance trades off the FOV in the $Y$ direction [Figs. 7(b) and 7(c)], because of the lateral shift from the large light sheet angle (Supplement 1). As the FOV varies with different working distances and light sheet angles, the potential configurations are summarized in Supplement 1, Figs. S14 and S15.
There are two general challenges in the design for mesoscopic OPM, determined by the low/moderate NA of the primary objective. The first is the limited axial resolution and hence the volumetric throughput. As described in Section 1, an additional optical element can be introduced to increase the light sheet angle and thus the axial resolution. Singh et al. [40] used an additional excitation path adopting the method of open-top LSM to achieve cellular resolution within a ${1.5}\;{\rm mm} \times {1.5\,\,\rm mm}$ FOV. However, the system complexity increased by the coalignment and synchronization of both excitation and imaging paths. We took a different approach by adding a transmission grating to diffract the light sheet and augment the excitation angle. The advantage is the simplicity without an additional optical path, achieving the scanning/descanning by the same galvanometer mirror. Obviously, using a higher-throughput primary objective (e.g., ${\rm NA} \geq {0.5}$, ${\le} {10x}$ magnification) would naturally improve the axial resolution and collect more light from the sample [39,40,52]. Yet, there still exists a second challenge of limited light collection efficiency, elaborated below.
The general OPM design abides a perfect imaging condition with a unity magnification (assuming air working media) between the object space and IIP [53]. However, this design would result in a dismal collection efficiency at the tertiary objective at ${\rm NA} \lt {0.5}$ since the angle of the oblique image is too small with respect to the primary optical axis [10]. To improve the collection efficiency, a diffraction grating [10] or a refractive fiber bundle imager [39,52] was placed at the IIP to redirect the light and achieved up to ${48}\% { \times T}$ light efficiency from the primary to the tertiary imaging system, where $T$ denotes the total transmission efficiency at all optical surfaces. Singh et al. [40] used an independent light sheet excitation path, so that the collection efficiency was maintained even under perfect imaging conditions. Here, we took a different approach by the diffractive light sheet and the demagnification design, to augment the angle of the oblique image [Eqs. (1) and (2)]. By that, we increased the efficiency at the tertiary objective to 80%, compensating for the loss at the grating. Using the off-the-shelf grating, the current overall efficiency from the primary to the tertiary imaging system is ${\sim}{16}\% { \times T}$ (i.e., 35% at the grating, 80% at the tertiary objective, 56% at the pupil mismatching). However, by customizing a polarization-selective transmission grating, the zeroth-order transmission can be improved from 35% to ${\sim}{50}\%$, and ${\gt}{90}\%$ in first-order diffraction for excitation. The optical magnification can also be carefully designed to eliminate pupil mismatching. These improvements would result in a light collection efficiency of ${\sim}{40}\% {\times T}$ (see Table S2), similar to that of typical OPM. A polarization-selective grating would also make the other orders of diffraction negligible, so that the moving aluminum block can be removed. To provide a comprehensive comparison, Table S2 (Supplement 1) summarizes the key specifications among several recent reports on large-scale OPM.
One intriguing observation is that our demagnification design does not significantly add aberrations in comparison with the perfect imaging condition [53]. We performed an additional Zemax simulation on two demagnified configurations with ${M_L} = {1/4}$, 1/2.5, and a perfect imaging condition with ${M_L} = {1}$. (The Zemax file for the simulation can be found in Dataset 1, Ref. [54], Dataset 2, Ref. [55], Dataset 3, Ref. [56], Dataset 4, Ref. [57], and Dataset 5, Ref. [58].) All three shared the identical ${10} \times$ primary objective lens (Supplement 1, Figs. S9–S12). We found that three configurations regardless of magnification can achieve diffraction-limited performance over ${\sim}{\pm} {1.25}\;{\rm mm}$ FOV (Supplement 1, Fig. S12). Considering that the field number of the ${10} \times$ primary objective is 26.5, it matched well with the simulation. We speculate that the demagnification design may not introduce significant aberration with a low/moderate NA primary objective, which certainly requires a more thorough examination on the applicable conditions.
We should note that the transmission grating is inherently wavelength dependent, which is inconvenient for simultaneous multicolor imaging and multiphoton implementation due to the dispersion. However, corrective optical design in the upstream optical path, for example, using a refractive prism, may be implemented to counter the dispersion caused by the grating, such that excitation beams with different wavelengths can be focused on the same position. In spite of the limitations, the simple addition of a transmission grating in Meso-OPM provides a straightforward method to obtain 3D cellular resolution over a mesoscopic FOV, a significant capability for observing 4D dynamics in a large volume as demonstrated herein.
As for future work, we will explore primary objective lenses with different NAs to explore combinations of FOV and resolution. Considering that the diffraction efficiency can be optimized, the proposed method only requires dry objectives without limitations on the NA, as long as the working distance is sufficient to accommodate the additional components. The lateral resolution will still follow Abbe’s diffraction limit while the axial resolution will maintain as it mainly depends on the diffractive angle introduced by the grating.
In summary, we circumvent the theoretical limitation of insufficient axial resolution in a low NA objective lens by creating a high-angle diffractive light sheet. By avoiding the trade-offs between FOV, axial resolution, and imaging speed, we demonstrate 4D cellular resolution over a FOV that was unattainable previously.
Funding
BrightFocus Foundation (2018132); National Institutes of Health (R01EY032163, R01NS108464, R01CA232015, F31EY032790, R01DC017785, and R01DC009607).
Disclosures
All other authors declare they have no competing interests. We acknowledge the Ahrens Lab of the Howard Hughes Medical Institute’s Janelia Research Campus for providing the jGCaMP7s expressing Zebrafish larvae.
Data availability
Data underlying the results presented in this paper are available in Dataset 1, Ref. [54], Dataset 2, Ref. [55], Dataset 3, Ref. [56], Dataset 4, Ref. [57], and Dataset 5, Ref. [58]. Raw images underlying the data are not publicly available at this time but may be obtained from the authors upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
REFERENCES
1. L. Tian, S. A. Hires, T. Mao, D. Huber, M. E. Chiappe, S. H. Chalasani, L. Petreanu, J. Akerboom, S. A. McKinney, E. R. Schreiter, C. I. Bargmann, V. Jayaraman, K. Svoboda, and L. L. Looger, “Imaging neural activity in worms, flies and mice with improved GCaMP calcium indicators,” Nat. Methods 6, 875–881 (2009). [CrossRef]
2. T.-W. Chen, T. J. Wardill, Y. Sun, S. R. Pulver, S. L. Renninger, A. Baohan, E. R. Schreiter, R. A. Kerr, M. B. Orger, V. Jayaraman, L. L. Looger, K. Svoboda, and D. S. Kim, “Ultrasensitive fluorescent proteins for imaging neuronal activity,” Nature 499, 295–300 (2013). [CrossRef]
3. J.-D. Pédelacq, S. Cabantous, T. Tran, T. C. Terwilliger, and G. S. Waldo, “Engineering and characterization of a superfolder green fluorescent protein,” Nat. Biotechnol. 24, 79–88 (2006). [CrossRef]
4. N. C. Shaner, R. E. Campbell, P. A. Steinbach, B. N. G. Giepmans, A. E. Palmer, and R. Y. Tsien, “Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein,” Nat. Biotechnol. 22, 1567–1572 (2004). [CrossRef]
5. R. K. Chhetri, F. Amat, Y. Wan, B. Höckendorf, W. C. Lemon, and P. J. Keller, “Whole-animal functional and developmental imaging with isotropic spatial resolution,” Nat. Methods 12, 1171–1178 (2015). [CrossRef]
6. A. Hernández, V. Nácher, R. Luna, A. Zainos, L. Lemus, M. Alvarez, Y. Vázquez, L. Camarillo, and R. Romo, “Decoding a perceptual decision process across cortex,” Neuron 66, 300–314 (2010). [CrossRef]
7. Z. V. Guo, N. Li, D. Huber, E. Ophir, D. Gutnisky, J. T. Ting, G. Feng, and K. Svoboda, “Flow of cortical activity underlying a tactile decision in mice,” Neuron 81, 179–194 (2014). [CrossRef]
8. P. S. Tsai, C. Mateo, J. J. Field, C. B. Schaffer, M. E. Anderson, and D. Kleinfeld, “Ultra-large field-of-view two-photon microscopy,” Opt. Express 23, 13833–13847 (2015). [CrossRef]
9. J. Fan, J. Suo, J. Wu, H. Xie, Y. Shen, F. Chen, G. Wang, L. Cao, G. Jin, Q. He, T. Li, G. Luan, L. Kong, Z. Zheng, and Q. Dai, “Video-rate imaging of biological dynamics at centimetre scale and micrometre resolution,” Nat. Photonics 13, 809–816 (2019). [CrossRef]
10. M. Hoffmann and B. Judkewitz, “Diffractive oblique plane microscopy,” Optica 6, 1166–1170 (2019). [CrossRef]
11. J. N. Stirman, I. T. Smith, M. W. Kudenov, and S. L. Smith, “Wide field-of-view, multi-region, two-photon imaging of neuronal activity in the mammalian brain,” Nat. Biotechnol. 34, 857–862 (2016). [CrossRef]
12. M. Clough, I. A. Chen, S.-W. Park, A. M. Ahrens, J. N. Stirman, S. L. Smith, and J. L. Chen, “Flexible simultaneous mesoscale two-photon imaging of neural activity at high speeds,” Nat. Commun. 12, 6638 (2021). [CrossRef]
13. W. Shao, K. Kilic, W. Yin, G. Wirak, X. Qin, H. Feng, D. Boas, C. V. Gabel, and J. Yi, “Wide field-of-view volumetric imaging by a mesoscopic scanning oblique plane microscopy with switchable objective lenses,” Quant. Imaging Med. Surg. 11, 98397–98997 (2021). [CrossRef]
14. Y. Xue, I. G. Davison, D. A. Boas, and L. Tian, “Single-shot 3D wide-field fluorescence imaging with a computational miniature mesoscope,” Sci. Adv. 6, eabb7508 (2020). [CrossRef]
15. G. McConnell, J. Trägårdh, R. Amor, J. Dempster, E. Reid, and W. B. Amos, “A novel optical microscope for imaging large embryos and tissue volumes with sub-cellular resolution throughout,” eLife 5, e18659 (2016). [CrossRef]
16. N. J. Sofroniew, D. Flickinger, J. King, and K. Svoboda, “A large field of view two-photon mesoscope with subcellular resolution for in vivo imaging,” eLife 5, e14472 (2016). [CrossRef]
17. A. W. Lohmann, R. G. Dorsch, D. Mendlovic, Z. Zalevsky, and C. Ferreira, “Space–bandwidth product of optical signals and systems,” J. Opt. Soc. Am. A 13, 470–473 (1996). [CrossRef]
18. Z. Zhang, L. Bai, L. Cong, P. Yu, T. Zhang, W. Shi, F. Li, J. Du, and K. Wang, “Imaging volumetric dynamics at high speed in mouse and zebrafish brain with confocal light field microscopy,” Nat. Biotechnol. 39, 74–83 (2021). [CrossRef]
19. V. Voleti, K. B. Patel, W. Li, C. P. 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, 1054–1062 (2019). [CrossRef]
20. T. Nöbauer, O. Skocek, A. J. Pernía-Andrade, L. Weilguny, F. M. Traub, M. I. Molodtsov, and A. Vaziri, “Video rate volumetric Ca2+ imaging across cortex using seeded iterative demixing (SID) microscopy,” Nat. Methods 14, 811–818 (2017). [CrossRef]
21. M. Levoy, R. Ng, A. Adams, M. Footer, and M. Horowitz, “Light field microscopy,” ACM Trans. Graph. 25, 924–934 (2006). [CrossRef]
22. R. Prevedel, Y.-G. Yoon, M. Hoffmann, N. Pak, G. Wetzstein, S. Kato, T. Schrödel, R. Raskar, M. Zimmer, E. S. Boyden, and A. Vaziri, “Simultaneous whole-animal 3D imaging of neuronal activity using light-field microscopy,” Nat. Methods 11, 727–730 (2014). [CrossRef]
23. K. Wang, D. E. Milkie, A. Saxena, P. Engerer, T. Misgeld, M. E. Bronner, J. Mumm, and E. Betzig, “Rapid adaptive optical recovery of optimal resolution over large volumes,” Nat. Methods 11, 625–628 (2014). [CrossRef]
24. 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, 1065–1069 (2008). [CrossRef]
25. M. B. Ahrens, M. B. Orger, D. N. Robson, J. M. Li, and P. J. Keller, “Whole-brain functional imaging at cellular resolution using light-sheet microscopy,” Nat. Methods 10, 413–420 (2013). [CrossRef]
26. J. Huisken, J. Swoger, F. D. Bene, J. Wittbrodt, and E. H. K. Stelzer, “Optical sectioning deep inside live embryos by selective plane illumination microscopy,” Science 305, 1007–1009 (2004). [CrossRef]
27. A. Kumar, Y. Wu, R. Christensen, P. Chandris, W. Gandler, E. McCreedy, A. Bokinsky, D. A. Colón-Ramos, Z. Bao, M. McAuliffe, G. Rondeau, and H. Shroff, “Dual-view plane illumination microscopy for rapid and spatially isotropic imaging,” Nat. Protoc. 9, 2555–2573 (2014). [CrossRef]
28. B.-C. Chen, W. R. Legant, K. Wang, et al., “Lattice light-sheet microscopy: imaging molecules to embryos at high spatiotemporal resolution,” Science 346, 1257998 (2014). [CrossRef]
29. T.-L. Liu, S. Upadhyayula, D. E. Milkie, et al., “Observing the cell in its native state: Imaging subcellular dynamics in multicellular organisms,” Science 360, eaaq1392 (2018). [CrossRef]
30. B. Migliori, M. S. Datta, C. Dupre, M. C. Apak, S. Asano, R. Gao, E. S. Boyden, O. Hermanson, R. Yuste, and R. Tomer, “Light sheet theta microscopy for rapid high-resolution imaging of large biological samples,” BMC Biol. 16, 57 (2018). [CrossRef]
31. A. K. Glaser, N. P. Reder, Y. Chen, C. Yin, L. Wei, S. Kang, L. A. Barner, W. Xie, E. F. McCarty, C. Mao, A. R. Halpern, C. R. Stoltzfus, J. S. Daniels, M. Y. Gerner, P. R. Nicovich, J. C. Vaughan, L. D. True, and J. T. C. Liu, “Multi-immersion open-top light-sheet microscope for high-throughput imaging of cleared tissues,” Nat. Commun. 10, 2781 (2019). [CrossRef]
32. C. Dunsby, “Optically sectioned imaging by oblique plane microscopy,” Opt. Express 16, 20306–20316 (2008). [CrossRef]
33. 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, 113–119 (2015). [CrossRef]
34. M. Kumar, S. Kishore, J. Nasenbeny, D. L. McLean, and Y. Kozorovitskiy, “Integrated one- and two-photon scanned oblique plane illumination (SOPi) microscopy for rapid volumetric imaging,” Opt. Express 26, 13027–13041 (2018). [CrossRef]
35. H. Sparks, L. Dent, C. Bakal, A. Behrens, G. Salbreux, C. Dunsby, and C. Dunsby, “Dual-view oblique plane microscopy (dOPM),” Biomed. Opt. Express 11, 7204–7220 (2020). [CrossRef]
36. J. Kim, M. Wojcik, Y. Wang, S. Moon, E. A. Zin, N. Marnani, Z. L. Newman, J. G. Flannery, K. Xu, and X. Zhang, “Oblique-plane single-molecule localization microscopy for tissues and small intact animals,” Nat. Methods 16, 853–857 (2019). [CrossRef]
37. B. Yang, X. Chen, Y. Wang, S. Feng, V. Pessino, N. Stuurman, N. H. Cho, K. W. Cheng, S. J. Lord, L. Xu, D. Xie, R. D. Mullins, M. D. Leonetti, and B. Huang, “Epi-illumination SPIM for volumetric imaging with high spatial-temporal resolution,” Nat. Methods 16, 501–504 (2019). [CrossRef]
38. E. Sapoznik, B.-J. Chang, J. Huh, et al., “A versatile oblique plane microscope for large-scale and high-resolution imaging of subcellular dynamics,” eLife 9, e57681 (2020). [CrossRef]
39. M. Hoffmann, J. Henninger, L. Richter, and B. Judkewitz, “Brain-wide imaging of an adult vertebrate with image transfer oblique plane microscopy,” bioRxiv2022.05.16.492103 (2022). [CrossRef]
40. R. Singh, K. Subramanian, R. M. Power, A. Paix, A. Ikmi, and R. Prevedel, “An oblique plane microscope for mesoscopic imaging of freely moving organisms with cellular resolution,” bioRxiv2022.07.15.500249 (2022). [CrossRef]
41. M. Kumar and Y. Kozorovitskiy, “Tilt (in)variant lateral scan in oblique plane microscopy: a geometrical optics approach,” Biomed. Opt. Express 11, 3346–3359 (2020). [CrossRef]
42. L. Zhang, A. Capilla, W. Song, G. Mostoslavsky, and J. Yi, “Oblique scanning laser microscopy for simultaneously volumetric structural and molecular imaging using only one raster scan,” Sci. Rep. 7, 1 (2017). [CrossRef]
43. W. Shao, W. Shao, J. Yi, and J. Yi, “Non-interferometric volumetric imaging in living human retina by confocal oblique scanning laser ophthalmoscopy,” Biomed. Opt. Express 13, 3576–3592 (2022). [CrossRef]
44. S. J. Cruikshank, H. J. Rose, and R. Metherate, “Auditory thalamocortical synaptic transmission in vitro,” DJ. Neurophysiol. 87, 361–384 (2002). [CrossRef]
45. C. B. Kimmel, W. W. Ballard, S. R. Kimmel, B. Ullmann, and T. F. Schilling, “Stages of embryonic development of the zebrafish,” Dev. Dyn 203, 253–310 (1995). [CrossRef]
46. R. Kelsh, M. Brand, Y. J. Jiang, C. Heisenberg, S. Lin, P. Haffter, J. Odenthal, M. Mullins, F. V. van Eeden, M. Furutani-Seiki, M. Granato, M. Hammerschmidt, D. Kane, R. M. Warga, D. Beuchle, L. Vogelsang, and C. Nüsslein-Volhard, “Zebrafish pigmentation mutations and the processes of neural crest development,” Development 123, 369–389 (1996). [CrossRef]
47. H. Dana, Y. Sun, B. Mohar, B. K. Hulse, A. M. Kerlin, J. P. Hasseman, G. Tsegaye, A. Tsang, A. Wong, R. Patel, J. J. Macklin, Y. Chen, A. Konnerth, V. Jayaraman, L. L. Looger, E. R. Schreiter, K. Svoboda, and D. S. Kim, “High-performance calcium sensors for imaging activity in neuronal populations and microcompartments,” Nat. Methods 16, 649–657 (2019). [CrossRef]
48. O. A. Shemesh, C. Linghu, K. D. Piatkevich, et al., “Precision calcium imaging of dense neural populations via a cell-body-targeted calcium indicator,” Neuron 107, 470–486.e11 (2020). [CrossRef]
49. O. Randlett, C. L. Wee, E. A. Naumann, O. Nnaemeka, D. Schoppik, J. E. Fitzgerald, R. Portugues, A. M. B. Lacoste, C. Riegler, F. Engert, and A. F. Schier, “Whole-brain activity mapping onto a zebrafish brain atlas,” Nat. Methods 12, 1039–1046 (2015). [CrossRef]
50. M. Kunst, E. Laurell, N. Mokayes, A. Kramer, F. Kubo, A. M. Fernandes, D. Förster, M. Dal Maschio, and H. Baier, “A cellular-resolution atlas of the larval zebrafish brain,” Neuron 103, 21 (2019). [CrossRef]
51. K. Baker, K. S. Warren, G. Yellen, and M. C. Fishman, “Defective ‘pacemaker’ current (Ih) in a zebrafish mutant with a slow heart rate,” Proc. Natl. Acad. Sci. USA 94, 4554–4559 (1997). [CrossRef]
52. G. S. Lee, W. Liang, and E. M. C. Hillman, “MesoSCAPE—highspeed functional volumetric imaging of multi-millimeter biological sample with cellular resolution,” in Biophotonics Congress: Biomedical Optics 2022 (Translational, Microscopy, OCT, OTS, BRAIN) (Optica Publishing Group, 2022), paper BM4C.5.
53. E. J. Botcherby, R. Juškaitis, M. J. Booth, and T. Wilson, “An optical technique for remote focusing in microscopy,” Opt. Commun. 281, 880–887 (2008). [CrossRef]
54. J. Yi, “Zemax simulation with immersion media,” figshare (2022), https://doi.org/10.6084/m9.figshare.21282516.
55. J. Yi, “Zemax simulation without immersion media,” figshare (2022), https://doi.org/10.6084/m9.figshare.21282600.
56. J. Yi, “Zemax simulation when magnification is 1,” figshare (2022), https://doi.org/10.6084/m9.figshare.21282612.
57. J. Yi, “Zemax simulation when magnification is 1/2.5,” figshare (2022), https://doi.org/10.6084/m9.figshare.21282609.
58. J. Yi, “Zemax simulation when magnification is 1/4,” figshare (2022), https://doi.org/10.6084/m9.figshare.21282615.
