Open Access
Add to BibSonomy
Abstract
Multifocal illumination can improve image acquisition time compared to single point scanning in confocal microscopy. However, due to an increase in the system complexity, obtaining uniform multifocal illumination throughout the field of view with conventional methods is challenging. Here, we propose a volume holographic lenslet array illuminator (VHLAI) for multifocal confocal microscopy. To obtain uniform array illumination, a super Gaussian (SG) beam has been incorporated through VHLAI with an efficiency of 43%, and implemented in a confocal microscope. The design method for a photo-polymer based volume holographic beam shaper is presented and its advantages are thoroughly addressed. The proposed system can significantly improve image acquisition time without sacrificing the quality of the image. The performance of the proposed multifocal confocal microscopy was compared with wide-field images and also evaluated by measuring optically sectioned microscopic images of fluorescence beads, florescence pollen grains, and biological samples. The proposed multifocal confocal system generates images faster without any changes in scanning devices. The present method may find important applications in high-speed multifocal microscopy platforms.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Confocal laser scanning microscopy is one of the standard imaging modalities for industrial and biomedical imaging applications [1,2]. The key feature of confocal microscopy is optical sectioning capability, which helps in obtaining high resolution and high contrast images of volumetric samples [3,4]. A laser scanning-based confocal microscope can be used for both label-free and fluorescence imaging in reflection and transmission modes [5,6]. Keeping the conventional method as a standard, many advanced functionalities have been added to the confocal microscope for specific advantages such as higher spatial and temporal resolutions, higher SNR, and imaging for particular applications [7,8]. Common single focal point scanning in confocal microscope imposes limitations in controlling the exposure and scanning time, which restricts its range of applications [9,10].
To augment imaging speed, several parallel image acquisition methods have been implemented [11]. In that regard, multifocal illumination and detection can be a perfect way to reduce the image acquisition time [12]. To reduce the scanning time, the multifocal spots in the sample plane can be obtained by focusing on the beam array [13]. Laterally or axially separated multifocal points for imaging have been generated by various methods [14,15]. Generation of multifocal illumination in the single plane, as well as multi-plane, has been reported [16]. In multifocal confocal microscopy, image stitching methods have been used to reconstruct the overall image with reduced scanning time [17]. The use of Nipkow disk scanning limits the accuracy because of the vibration of the rotating disk, which affects its illumination efficiency [6]. Similar to the spinning disk, to increase the imaging speed, various multifocal illumination techniques such as pinhole arrays, slit apertures, liquid crystal display (LCD) panels, and micro-electro-mechanical systems (MEMS) have been implemented in confocal systems [18,19]. Factors like complex alignment, lack of efficiency, cost, and complexity of the system limit the range of applications [18].
One of the simple and direct techniques for generating multifocal spots is made by illuminating a microlens array (MLA) with a Gaussian beam [20]. However, due to its inhomogeneous amplitude distribution from center to the periphery, peripheral points receive less intensity as compared to center over the entire field of view (FOV) [21,22]. There are many techniques to produce homogeneous multifocal illumination, such as diffractive optical elements (DOEs), spatial light modulators (SLMs), multimode fibers, beam splitters, and MEMS [21]. However, each of these techniques is complex and required multiple components and operations [22]. Alternatively, illuminating the multifocal array with an SG beam provides an array of multifocal spots with a homogeneous intensity distribution.
Digital micro-mirror device (DMD), spatial light modulators (SLM), and beam shaper have commonly been used to shape the Gaussian beam into super-Gaussian (SG) beam [23]. Implementation of such a device for illumination in a confocal microscope increases the system’s size and cost. Volume holographic optical element is one of the potential alternatives to reduce the complexity by replacing these devices with additional advantages. Previous studies successfully exhibited the generation of axial multifocal spots and various complex beam shapes incorporated with volume holographic optical elements [15,24,25]. The advantages of volume holographic beam shapers include compact size, high angular selectivity, and high wavelength degeneracy for solving the above issues [26,27,28]. Here we chose PQ: PMMA photopolymer material to make a volume holography-based lenslet array illuminator. Because of its excellent properties such, as high diffraction efficiencies, lightweight, low amount of diminution, simple fabrication procedure, make photopolymer-based volume hologram a unique candidate [15,29,30]. By incorporating multifocal illumination components in a confocal microscope, a physical pinhole can be replaced with a virtual pinhole in the detection arm [31,32]. Various research works prominently demonstrate the use of CCD or CMOS cameras instead of pinhole-APD pairs for their distinct imaging applications [32,33]. Image moment analysis, Gaussian fit, or subtraction imaging algorithm techniques were widely used in this situation for image acquisition and processing [34,35].
In this work, we designed and fabricated a VHLAI which provides multifocal flattop illumination for confocal microscopy. Instead of a regular lenslet array, we fabricated a volume holographic lenslet array illumination with embedded super-Gaussian amplitude distribution, multifocal spot generation with uniform intensity distribution. Our VHLAI is multipurpose and it can also be implemented in other imaging modalities. We demonstrate the unique advantages of VHLAI in confocal microscopy. The proposed confocal system with multifocal illumination enhances image acquisition time significantly when compared to single focal scanning confocal microscopy. A CCD camera with digital pinholes was used to detect signals and the images were generated with the help of an image stitching algorithm. Imaging characteristics such as lateral, axial resolution, and optical sectioning capability were experimentally measured. The imaging performance of our proposed system was evaluated through fluorescence imaging of fluorescence beads and fluorescence pollen grains. To demonstrate the volumetric imaging capability of our system for biological samples, fluorescent-labeled mouse cardiac tissues have been imaged for two depths, and a significant out-of-focus background suppression has been obtained with respect to corresponding wide-field image. The major advantage of the system is the improvement in image acquisition time without sacrificing spatial resolutions.
2. Volume holographic gratings
2.1 Properties of volume holograms
Volume holographic gratings offer multiple unique diffractive properties as compared to thin gratings [24]. Diffraction from thin gratings occurs in the Raman-Nath regime and it produces multiple diffraction orders, whereas diffraction from volume holographic grating occurs in Bragg's regime with dominant single diffraction order [27]. Apart from the application of volume holographic gratings in imaging, it has also been used as an illumination element in microscopy [14,24,26]. Here, we fabricated photopolymer-based volume holographic grating corresponding to multifocal SG lenslet array. To record a volume holographic grating, an interference pattern between the signal and reference beam is created on PQ: PMMA photopolymer. The refractive index modulation corresponding to the grating formation inside the photosensitive material can be written as
where n0 is the average refractive index inside the material and n1 small refractive index modulation within the material. The grating vector ${\vec{K}_G}$, and the wave vectors of the reference ${\vec{K}_R}$ and the signal ${\vec{K}_S}$ wave is given by ${\theta _R}$ and ${\theta _S}$ are the angle of incidence for the reference and the signal beams, $k = 2\pi /\lambda $ is the wave-number, $\lambda $ is the operation wavelength. For our present work we set the recording angle for reference and signal beams as $({{\theta_R} = {{34}^0},\; and\; {\theta_s} ={-} {{34}^0}\; } )$. The expression for the Bragg wavelength degeneracy between recording and reconstruction beam can be given as [27]The period of the volume grating can be calculated as
where $\Delta \theta $ is the angular shift of the probe beam with operation wavelength (λ0+Δλ) from the Bragg-matched angle θR of the recording beam. $\phi $ represents slant angle of the grating vectors, n is the refractive index value of the photosensitive medium. The multi-wavelength reconstruction of volume holographic grating is possible due to the Bragg wavelength degeneracy property. The multi-wavelength operation of volume holograms can be described by the K-sphere diagram [24–29]. A change in the wavelength of the reconstructed light results in the corresponding change in the diffraction angle and the Bragg matched condition is satisfied [27–29].2.2 Super Gaussian beam shaping
Gaussian beams have been restricted from many laser applications because of their inhomogeneous nature their beam radius [36]. To overcome this problem an SG beam has been proposed, which has uniform intensity distribution all over its diameter. Depending on the shape it has also been known as a flattop beam [37]. By redistribution of irradiance over beam cross-section, SG beams with homogeneous distribution all over its diameter can be obtained. For the target profile of an SG beam, cross-sectional intensity distribution can be described as,
where ${I_{SG}}$ is the intensity of the SG beam, ${I_{SG0}}$ is the intensity at the center of the beam waist, N ($N$ ≥ 2) is the beam order, and wb is the beam waist [38]. Figure 1 shows the simulation results of Gaussian and SG beams with different orders for circular and square shapes. In our experiment, a DMD is used to generate a super-Gaussian profile by reshaping the incident Gaussian beam by the Lee hologram method. The amplitude distribution of the SG beam is encoded on a Lee hologram using the following equation [38]
Fig. 1. Simulation results, Lee hologram, and experimental results of Gaussian, SG (circular), and SG (square) beam shapes are tabulated respectively, and corresponding intensity profiles from simulated and experimental results.
2.3 Volume holographic lenslet array illuminator
To obtain a VHLAI, a volume holographic recording setup capable of precise control over its angular and translational movements is required [29]. Figure 2 shows the illustration of the recording setup for VHLAI on a thick PQ: PMMA substrate. The chemical process for preparing PQ: PMMA photopolymer is carried out in accordance with the previous works [29]. Free Radicle polymerization process with weight ratio of 100:0.5:0.7 is used to prepare substrate. The MMA in liquid form is mixed with powdered PQ and AIBN at a constant temperature of 40°C. After filtering the solution, the mixture is filled inside the mold and performed curing at temperature to 50°C for 120 hours, which results in the solidified photosensitive polymer substrate. A blue laser (Innova 304C, Coherent Inc.), at a wavelength of λ=488 nm, is used as the light source in the interferometer. In the signal beam path, a DMD (DLP Light Crafter 6500) is used to shape the beam. A Lee hologram encoded with the amplitude of SG beam is projected on DMD. The generated SG beam is then passed through a Micro Lens Array (Thorlabs MLA150-5C) made up of fused silica substrate which comprises an array of plano-convex lenses within a 10mm×10mm square area to convert a single beam into multiple beam spots. The size of the individual focal spot is near the diffraction limit.
Fig. 2. Illustration of the experimental setup for the recording of VHLAI on PQ: PMMA substrate. The signal arm of the system consists of a microlens array, MLA (focal length, f = 5.2 mm), and relay lenses, L1 and L2 (focal length, f = 50 mm) to generate an array of uniform multifocal points.
The photopolymer substrate is kept before the multifocal points in the illumination path and a reference Gaussian beam is allowed to interfere with the multiple beam spots at a recording angle (θr) of 34°. The exposure time and the incident beam power for the two-beam are 21 seconds and 5.6 mW respectively. Figure 3(a) shows the recorded and reconstruction results of uniform focal spots using DMD and VHLAI respectively with corresponding intensity profiles. In our experiment we used an illumination beam with an intensity 0.525 w/cm2 for recording and reconstruction of holograms and imaging using VHLAI. It is evident that intensity profiles of a reconstructed array of focal spots are having equal intensity distribution all over the field of view.
Fig. 3. (a) The intensity distribution of multifocal spots generated from the DMD and VHLAI, respectively. (b) Corresponding normalized intensity profiles along the dashed red lines. (c) Experimental and simulated angular selectivity curve of VHLAI. The FWHM of the angular selectivity curve is ∼ 0.06°.
To obtain the angular selectivity curve variation of diffracted power of the reconstructed beam is measured with respect to rotation angle (Δθ). Figure 3(c) depicts the angular selectivity curve of VHLAI. The maximum diffraction efficiency of the focal spots is around 43% and the FWHM is ∼ 0.060 which shows high angular selectivity of VHLAI. The experimentally obtained values of diffraction efficiency are denoted by black dots and corresponding simulation results are plotted with a solid black line curve. These results show the inherent high angular selectivity features of our volume holographic element. Diffraction efficiency is calculated using rigorous coupled-wave theory [26].
The parameters for PQ: PMMA in our calculations are (thickness= 1.8 mm, refractive index n = 1.49, absorption co-efficient = 0.009/mm). From Fig. 3(c) it is evident that the FWHM of angular selectivity curve from experimental and simulated results are in good agreement. PQ: PMMA photopolymer based volume holographic gratings are essentially phase gratings and can achieve very high diffraction efficiency of up to 80% by optimizing fabrication and the recording parameters [29]. The multifocal spot generation from volume hologram still has room for improvement in efficiency. By adapting pre exposure and the dark delay time, higher efficiencies of the volume holographic gratings can be achieved [28].
3. Multifocal confocal imaging system using VHLAI
A multifocal confocal microscope is constructed by implementing a VHLAI in the illumination path of the system to obtain multifocal illumination at the sample plane. Figure 4 shows the schematic diagram of a multifocal confocal microscope with the VHLAI. A blue laser (Innova 304C, Coherent Inc.) at λ=488 nm is used as a light source. The beam is expanded to illuminate VHLAI using lenses, L1 and L2. The diffraction pattern from VHLAI generates uniformly intensity multifocal points with super-Gaussian intensity distribution as shown at the bottom of Fig. 4.
Fig. 4. Schematic diagram of the multifocal confocal microscope system. A (9×9) array of SG focal spots is shown at the bottom. A VHLAI is used as an illuminator in the illumination path. Lens L1 and L2 are used to construct a beam expander. Lens L3 is used to project diffraction patterns from VHLAI to the back focal plane of the objective lens, OL. A polarizer, P - quarter-wave plate, QWP pair is used to control the polarization of the illumination. A beam splitter, BS (50:50) used to split excitation and emission light. S denotes the sample. Lens L4 is used as an imaging lens. A fluorescence filter is used to separate background light in the emission path and a CCD to capture the signal.
The multifocal spot with SG intensity distribution from the VHLAI is projected onto the back focal plane of the microscopic objective lens using a relay lens, L3 (focal length, f = 100 mm). A polarizer (P) and a quarter-wave plate (QWP) are used in the illumination path to control the polarization. A beam splitter, BS (Thorlabs, 50:50) is placed in the illumination beam path. An objective lens, OL (Olympus, 50X, numerical aperture NA= 0.55, focal length, f = 2 mm) is used to illuminate the sample with multifocal points. The reflected light from the sample is collected by the microscopic objective and guided to the detection arm. The light reflected from beam splitter is imaged through lens L4 (fl = 150 mm) and a fluorescence filter (Chroma, band-pass, CWL = 535 nm, FWHM = 30 nm) onto the CCD camera (Canon EOS 650D). The sample is stage scanned with a stepper motor-based scanning system (Thorlabs NanoMax-TS 341/M). A customized program has been made for stage scanning and data acquisition using LabVIEW 2014. An exposure time of 1/250 seconds has been coded. For CCD, the frequency was set at 30 frames per second (fps). The setting of the camera frequency can greatly influence the scanning time.
4. Imaging using a multifocal confocal microscope with VHLAI
4.1 Selection of scanning step size and digital pinhole size
The image acquisition speed in a multifocal confocal system is decided by the number of the lenslet, their separation, camera frame rate, field of view, and speed of scanning device. We used 81 focal spots as an example, and the number of focal spots can be further increased according to the need and specifications of hardware used. The number of lenslet in the illumination path is decided by the objective lens's back pupil diameter. The lens array must perfectly fill the back pupil of the objective lens. To increase the number of focal spots in the illumination array, we need to reduce the distance between each lenslet while hologram recording. To avoid the crosstalk that can result in poor image quality, an equation is used to decide the distance between the adjacent focal spot generated by the lenslet. It depends on the distance between the lenslet in the sample and image planes, the focal lengths of the lenslet, objective lens, and tube lens, respectively, as shown in Eq. (9). Basic point-by-point stage scanning is done using a stepper motor-based scanning system. To measure the intensity distribution of focal spots, a mirror is used as a sample. The distance between each of the focal spots generated from VHLAI can be measured using the intensity profile analysis. The pitch between focal spots at the detector plane $({d_l})$ is 172 µm. Using the following equation, the distance between the focal spots at the sample plane $({d_s})$ can be calculated as [20]
where the focal length of the lens, L4 (${f_l} = 150$ mm) and focal length of the objective lens ${f_0}$ = 3 mm. Using a focal length of tube lens and objective lens the value of ${d_s}$ is obtained as 5.16 µm.Here we set the step size of scanning in stepper motor as 1µm. We scanned 200×200 steps in the lateral direction to collect enough data to provide overlap while stitching images. The minimum scanning step size criteria to fulfill the maximum optical resolution can be calculated using the Nyquist-Shannon sampling theorem [39]. The maximum spatial frequency that can be measured by the current system is group 8 element 6 (i.e., 1.096 µm). The 1 Airy Unit (AU) for our system is around 20.025 µm, which is obtained by measuring focal spot size at the detection plane. The pixel size of CCD is 4.3 µm, for our case 5×5 pixels corresponds to 1AU digital pinhole. To achieve a good confocality, we choose a digital pinhole in AU between 0.5AU-0.75AU. We considered 3×3 pixels in CCD which corresponds to a digital pinhole size of 0.64AU (physical size =12.9 µm).
4.2 Image reconstruction
The signal reconstruction is done by the image moments method which can be explained by the following equation [33]
$I({x,y} )$ is the intensity at each pixel location $({x,y} )$ and i, j = 0, 1, 2, … m. The summation is carried out over 3×3 pixels’ area according to digital pinhole selection. The irradiance is calculated for each point by measuring the zeroth-order (m = 0) image moment ${M_{00}}.$ It has been implemented with a home-built system automation code using LabVIEW software.Combining the signals from each focal spot after scanning may result in boundary discontinuities. To avoid this problem a normalization process is needed before stitching [40]. A highly reflective mirror was used as a sample to measure the intensity normalization constant. The image artifacts can be reduced by comparing and normalizing the edge pixel values of two neighboring component images. To make the whole image continuous, we chose an overlap of 3 pixels on each edge of component images. Average values are taken while overlapping the pixels on each edge of component images. We used Matlab Software for performing the stitching process.
5. Experimental results
To evaluate the lateral resolution of our multifocal confocal microscope a high-resolution USAF test target (positive type, Newport) was imaged as shown in Fig. 5(a). When the chart is brought into focus, we can resolve a line width of 1.096 µm (group 8 element 6) as shown in Fig. 5(b). Line profiles have been drawn for horizontal and vertical bars to calculate the contrast values, as shown in Fig. 5(c)–5(d). The optical sectioning capability of the system is experimentally evaluated by axially scanning the fluorescently labeled one-micrometer-sized green microsphere (Polyscience). Since we have multiple focal spots in the field of view, we measured the axial point spread function corresponding to four different focal spots and calculated their average value as shown in Fig. 5(e). From the graph, it is clear that axial resolution for individual focal spots is identical. The full width at half maximum (FWHM) of the axial curve is compared with the theoretical value which can be calculated using the following equation [8]
where $\lambda $ is the wavelength of excitation which is 488 nm, n is the refractive index of the air taken as unity, and the numerical aperture ($NA$) of the objective lens is 0.55. The theoretical value of FWHM in axial resolution analysis is 2.79 µm whereas the experimental value is 3.1 µm. The small deviations in the theoretical and experimental may be due to experimental measurements errors.Fig. 5. (a) Image of USAF 1951 high-resolution target. (b) Images of lowest resolved group element (group 8 element 6). (c)- (d) Corresponding line profile of lowest resolved group element (b). (e) Axial response curve plotted by axially scanning a mirror and averaged for different focal spots from the array.
To show the optical sectioning capability, we have carried out volumetric imaging of fluorescent-labeled green microspheres (25 µm in diameter, Polyscience) embedded in agarose (Invitrogen) as shown in Fig. 6. A wide-field image of green microspheres within the region of interest is shown in Fig. 6(a). Microspheres that are in-focus and out-of-focus are marked with red and yellow dashed circles, respectively. Figures 6(b)–6(c) show optically sectioned multifocal confocal images obtained by array illumination at z = 1.72 mm and z = 1.77 mm respectively.
Fig. 6. (a) Wide-field fluorescent image of green microspheres (25 µm in diameter, Polyscience), and corresponding optical sectioned multifocal confocal images at depths (b) z = 1.72 mm and, (c) z = 1.77 mm respectively. (d)- (e) Corresponding intensity profiles along red dotted line on (b) and (c) respectively. Scale bar denotes 30 µm.
From Fig. 6(b)–6(c) it is evident that the two depths show different in-focus microspheres. The corresponding intensity profiles plotted along the red-dash lines drawn in both layers are shown in Fig. 6(d)–6(e). Using the profiles, we compared the contrast variation between wide-field (plotted along the black dashed line) and confocal images from two different depths. The above results clearly show the capability of our multifocal confocal system in suppressing the out-of-focus background light for high-resolution volumetric imaging applications.
To illustrate the performance of our system for fluorescent imaging, a fluorescent-labeled pollen grain sample (Carolina biological supply company) is imaged. Figure 7(a) depicts a wide-field image of a pollen grain (Helianthus) with an average size of 60-95 µm. As we can see smaller spikes on the pollen grain surface are barely visible through wide-field microscopic images. Figures 7(b)–7(c) show corresponding optically sectioned multifocal confocal images at depths z = 1.73 mm and z = 1.75 mm respectively. Due to optical sectioning, different spikes on the pollen grain surface are visible for both depths. Figures 7(d)–7(e) show the intensity profiles of the corresponding blue dashed lines plotted on Fig. 7(b)–7(c). By analyzing the intensity profile from both layers, it can be observed that the tiny spikes that are pointed out on both layers with yellow arrows are separated as peaks in the profile. Whereas, the line profile of the wide-field image is progressing flat without any observable spikes or features. The differences in peaks are highlighted for comparison using gray rectangular strips. This measurement reveals the potential of the proposed system in fluorescence imaging to detect features of a sample that are minute and inseparable through wide-field imaging techniques.
Fig. 7. (a) A wide-field image of fluorescent-labeled pollen grain (Helianthus) with the size around 60-95 µm. Corresponding optical sectioned images obtained from multifocal confocal microscopy at depths (b) z = 1.73 mm and, (c) z = 1.75 mm respectively. (d)-(e) corresponding normalized intensity profiles with yellow arrows denoting the features from (b) and (c). Scale bar denotes 30 µm.
Confocal imaging of fluorescently tagged cardiac tissue is a reasonably fast way of 3-D visualization in straightforward and minimal-effort [41]. To demonstrate the fluorescence imaging capability of the multifocal confocal system for biomedical imaging, a fluorescent-labeled (Green fluorescent protein, GFP) mouse cardiac tissue is taken as a sample. Figures 8(a) and 8(g) show wide-field images of fluorescently labeled mouse cardiac tissue. There are no separable fine features visible in wide-field images. Figures 8(b) and 8(h) show corresponding optically sectioned multifocal confocal images of fluorescent-labeled mouse cardiac tissue obtained from depths of z = 1.11 mm and z = 1.02 mm, respectively. The minute features are visible after zooming into the white dashed box regions shown in Fig. 8(c)–8(d), 8(i)–8(j) from both layers. As shown in Fig. 8(e)–8(f), 8(k)–8(l), intensity profiles are plotted along the blue dashed lines drawn along the zoomed box areas from both layers. When examining the profiles from both layers, there are observable peaks through the plotted area that show discrete features from each layer within the area of interest. However, the profile of the wide-field image does not show any salient features. These images demonstrate the capability of the proposed confocal system for volumetric fluorescence imaging of biological samples [42].
Fig. 8. (a) A wide-field image of fluorescent-labeled mouse cardiac tissue, and (b) corresponding optical sectioned multifocal confocal images at a depth of z = 1.11 mm. (c)-(d) Zoomed in images of the dashed box region in (b). (e)-(f) Corresponding intensity profiles with yellow arrows denote features from (c) and (d). (g) A wide-field image of fluorescent-labeled mouse cardiac tissue, and (h) corresponding optical sectioned multifocal confocal images at a depth of z = 1.02 mm. (i)-(j) Zoomed in images of the dashed box region in (h). (k)- (l) Corresponding intensity profiles with yellow arrows denote features from (i) and (j). The gray-shaded region highlights the contrast variation in the profiles of both wide-field and confocal images. The scale bar denotes 30 µm.
Fig. 9. Comparison between different sections of fluorescent images from conventional confocal microscope and multifocal confocal microscope. (a) Optical sectioned confocal fluorescent image of fluorescent labeled mouse cardiac tissue taken by conventional confocal fluorescent microscope with a thickness around 8-10 µm. (b) Zoomed in image to dashed box region of (a). (c) Optical sectioned confocal fluorescent image of fluorescent labeled mouse cardiac tissue taken by proposed multifocal confocal fluorescence microscope. (d) Zoomed in image to dashed box region of (c). (e) Corresponding intensity profiles along the dashed line in both (b) and (d) respectively. (f) Corresponding histograms of both (b) and (d) respectively. Scale bar denotes 30 µm.
Here, we compared the quality of fluorescently labeled mouse heart tissue images obtained by the proposed multifocal confocal system with that of a conventional confocal microscope having identical components. The results of optical sectioned confocal fluorescent images as shown in Fig. 9. Figures 9(a) and 9(c) show images of fluorescently labeled mouse cardiac tissue. Corresponding zoom-in images from white dashed box regions are shown in Fig. 9(b) and 9(d). Intensity profile along the red and blue dashed lines are shown in Fig. 9(e). Histograms of images Fig. 9(b) and 9(d) are shown in Fig. 9(f) respectively. As we can see, the image quality of the two system are quite similar. A small difference is obtained due to the different sections of the sample taken at different time and experimental conditions.
Compared with the image acquisition time of conventional confocal scanning microscopy, the proposed multifocal confocal system can generate images significantly faster without any changes in scanning devices. An interesting observation from the experimental results is that the image quality was not sacrificed. In addition, the photo-bleaching rate is low, which has the potential for future studies in this area. This improved illumination technique can play important role in in-vivo and high-speed volumetric imaging applications. The volume holographic array illuminator designed and fabricated in this study may also find important applications in other microscopy modalities where multifocal spots with uniform illumination can enhance the imaging performance. The imaging speed of our system can be further improved by incorporating a large number of illumination focal spots and adapting a large CCD sensor size.
6. Discussion
In our multifocal confocal microscopic system, the resolution is decided by more than one factor. The theoretical lateral resolution of the objective lens (Olympus, 50X, NA = 0.55), which is used for our experiments, can be resolved up to 0.67 µm. However, another factor that limits the lateral resolution of our system is the scanning step size of the motorized stage. The lateral resolution reported in previous works [20] can reach up-to sub-micron levels by using piezo stage scanning with a step size of 50 nm. Also, to achieve better resolution and contrast, Gaussian fit methods have been adapted for post image processing. In our case, no post-processing methods are used for contrast improvement. There is a large scope for improvement in spatial resolution by implementing high resolution scanning devices such as piezo scanning stages or Galvano mirror scanners with finer step sizes that can provide diffraction limited imaging performance. Our system does not sacrifice lateral resolution at the cost of increasing scanning speed and image quality.
It should also be noted that it is a novel implementation of volume holographic optical elements in confocal microscopy through beam shaping concepts. Volume holographic optical elements can provide added advantages such as embedded uniform illumination, compact size, and ease of implementation, which can play a significant role in confocal microscope design. Our multifocal illumination can also be utilized in the other imaging modalities such as, endomicroscope, and light field imaging, as well. One of the most interesting and unique properties of a volume hologram is its multiplexing capability, which can provide advantages that thin gratings do not have or make difficult to obtain. Previous work demonstrated that wavelength coded volume holographic gratings can be recorded using the volume hologram's multiplexing property for multi-plane imaging [43]. The principle of wavelength multiplexing can be utilized to record a multi-color illumination pattern for multi-color confocal microscopy. A wavelength-coded volume holographic illuminator will be an efficient and direct means for multi-color fluorescence microscopy.
7. Conclusion
A photopolymer-based VHLAI is designed and fabricated using PQ: PMMA for generating multifocal spot array with uniform illumination for confocal microscopy. Images are generated from detected signals using a CCD camera and digital pinholes, with the help of image stitching algorithms. Our VHLAI can work for multiple wavelengths with high angular selectivity. The imaging characteristics of our system demonstrate that our multifocal confocal microscope can be implemented for biomedical imaging applications. The results demonstrate enhanced imaging speed with the significant suppression of out-of-focus fluorescence to achieve a lateral and axial resolution of 1.096 µm and 3.1 µm respectively (objective lens-50x). The illumination technique used in our confocal system is direct and more compact as compared to the other spatial light modulating devices. Multifocal illumination can significantly reduce scanning time without changing the scanning device with the added advantage of uniform illumination throughout the field of view. Our system is suitable for in-vivo and high-speed volumetric imaging for both biomedical and industrial applications.
Funding
National Taiwan University (08HZT49001, 108L7714, 109L7839, 111L894102); Ministry of Science and Technology, Taiwan (MOST 108-2221-E-002-168-MY4, MOST 108-2221-E-007-098-MY3).
Acknowledgments
The authors thank Ministry of Science and Technology (MOST), Taiwan for financial support.
Disclosures
The authors declare that there are no conflicts of interest related to this article.
Data availability
Data underlying the results presented in this paper is not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. M. Minsky, “Memoir on inventing the confocal scanning microscope,” Scanning 10(4), 128–138 (1988). [CrossRef]
2. J. B. Pawley and B. R. Masters, “Handbook of biological confocal microscopy,” Opt. Eng. 35(9), 2765–2766 (1996). [CrossRef]
3. J. A. Conchello and J. W. Lichtman, “Optical sectioning microscopy,” Nat. Methods 2(12), 920–931 (2005). [CrossRef]
4. D. Favro, R. L. Thomas, P.-K. Kuo, and L. Chen, “Confocal microscope,” U.S. patent 5,162,941 (1992).
5. R. H. Webb, “Confocal optical microscopy,” Rep. Prog. Phys. 59(3), 427–471 (1996). [CrossRef]
6. J. A. Conchello and J. W. Lichtman, “Theoretical analysis of a rotating-disk partially confocal scanning microscope,” Appl. Opt. 33(4), 585–596 (1994). [CrossRef]
7. M. Gu and C.J.R. Sheppard, “Confocal fluorescent microscopy with a finite-sized circular detector,” J. Opt. Soc. Am. A 9(1), 151–153 (1992). [CrossRef]
8. T. Wilson, “Resolution and optical sectioning in the confocal microscope,” J. Microsc. 244(2), 113–121 (2011). [CrossRef]
9. J. B. Pawley, “Fundamental and practical limits in confocal light microscopy,” Scanning 13(2), 184–198 (1991). [CrossRef]
10. J. Huff, “The Airy-scan detector from ZEISS: confocal imaging with improved signal-to-noise ratio and super-resolution,” Nat. Methods 12(12), i–ii (2015). [CrossRef]
11. H. J. Tiziani and H. M. Uhde, “Three-dimensional analysis by a microlens-array confocal arrangement,” Appl. Opt. 33(4), 567–572 (1994). [CrossRef]
12. T. Tanaami, S. Otsuki, N. Tomosada, Y. Kosugi, M. Shimizu, and H. Ishida, “High-speed 1-frame/ms scanning confocal microscope with a microlens and Nipkow disks,” Appl. Opt. 41(22), 4704–4708 (2002). [CrossRef]
13. W. Zhang, B. Yu, D. Y. Lin, H. H. Yu, S. W. Li, and J. L. Qu, “Optimized approach for optical sectioning enhancement in multifocal structured illumination microscopy,” Opt. Express 28(8), 10919–10927 (2020). [CrossRef]
14. P. H. Wang, V. R. Singh, J. M. Wong, K. B. Sung, and Y. Luo, “Non-axial-scanning multifocal confocal microscopy with multiplexed volume holographic gratings,” Opt. Lett. 42(2), 346–349 (2017). [CrossRef]
15. S. Vyas, Y. H. Chia, and Y. Luo, “Volume holographic spatial-spectral imaging systems [Invited],” J. Opt. Soc. Am. A 36(2), A47–A58 (2019). [CrossRef]
16. I. Khaw, B. Croop, J. Tang, A. Möhl, U. Fuchs, and K. Y. Han, “Flat-field illumination for quantitative fluorescence imaging,” Opt. Express 26(12), 15276–15288 (2018). [CrossRef]
17. J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch, S. Preibisch, C. Rueden, S. Saalfeld, B. Schmid, J. Y. Tinevez, D. J. White, V. Hartenstein, K. Eliceiri, P. Tomancak, and A. Cardona, “Fiji: an open-source platform for biological-image analysis,” Nat. Methods 9(7), 676–682 (2012). [CrossRef]
18. M. Ishihara and H. Sasaki, “High-speed surface measurement using a non-scanning multiple-beam confocal microscope,” Opt. Eng. 38(6), 1035–1040 (1999). [CrossRef]
19. S. Li and R. Liang, “DMD-based three-dimensional chromatic confocal microscopy,” Appl. Opt. 59(14), 4349–4356 (2020). [CrossRef]
20. Y. Yu, X. Ye, and M. D. McCluskey, “Confocal microscopy with a microlens array,” Appl. Opt. 59(10), 3058–3063 (2020). [CrossRef]
21. C. J. Rowlands, F. Ströhl, P. P.V. Ramirez, K. M. Scherer, and C. F. Kaminski, “Flat-Field Super-Resolution Localization Microscopy with a Low-Cost Refractive Beam-Shaping Element,” Sci. Rep. 8(1), 5630–5638 (2018). [CrossRef]
22. D. Mahecic, D. Gambarotto, K. M. Douglass, D. Fortun, N. Banterle, K. A. Ibrahim, M. Le Guennec, P. Gönczy, V. Hamel, P. Guichard, and S. Manley, “Homogeneous multifocal excitation for high-throughput super-resolution imaging,” Nat. Methods 17(7), 726–733 (2020). [CrossRef]
23. X. Ding, Y. Ren, and R. Lu, “Shaping super-Gaussian beam through digital micro-mirror device,” Sci. China Phys. Mech. Astron. 58(3), 1–6 (2015). [CrossRef]
24. S. Vyas, Y. H. Chia, and Y. Luo, “Conventional volume holography for unconventional Airy beam shapes,” Opt. Express 26(17), 21979–21990 (2018). [CrossRef]
25. T. Y. Hsieh, S. Vyas, J. C. Wu, and Y. Luo, “Volume holographic optical element for light sheet fluorescence microscopy,” Opt. Lett. 45(23), 6478–6481 (2020). [CrossRef]
26. H. Kogelnik, “Coupled Wave Theory for Thick Hologram Gratings,” Bell Syst. Tech. J. 48(9), 2909–2947 (1969). [CrossRef]
27. Y. Luo, J. Castro, J. K. Barton, R. K. Kostuk, and G. Barbastathis, “Simulations and experiments of aperiodic and multiplexed gratings in volume holographic imaging systems,” Opt. Express 18(18), 19273–19285 (2010). [CrossRef]
28. Y. Luo, J. M. Russo, R. K. Kostuk, and G. Barbastathis, “Silicon oxide nanoparticles doped PQ-PMMA for volume holographic imaging filters,” Opt. Lett. 35(8), 1269–1271 (2010). [CrossRef]
29. Y. Luo, P. J. Gelsinger, J. K. Barton, G. Barbastathis, and R. K. Kostuk, “Optimization of multiplexed holographic gratings in PQ-PMMA for spectral-spatial imaging filters,” Opt. Lett. 33(6), 566–568 (2008). [CrossRef]
30. T. Wilson and A. R. Carlini, “Three-dimensional imaging in confocal imaging systems with finite sized detectors,” J. Microsc. 149(1), 51–66 (1988). [CrossRef]
31. C. Yang and J. Mertz, “Transmission confocal laser scanning microscopy with a virtual pinhole based on nonlinear detection,” Opt. Lett. 28(4), 224–226 (2003). [CrossRef]
32. E. Sánchez-Ortiga, C. J. R. Sheppard, G. Saavedra, M. Martínez-Corral, A. Doblas, and A. Calatayud, “Subtractive imaging in confocal scanning microscopy using a CCD camera as a detector,” Opt. Lett. 37(7), 1280–1282 (2012). [CrossRef]
33. X. Ye and M. D. McCluskey, “Modular scanning confocal microscope with digital image processing,” PLoS One 11(11), e0166212 (2016). [CrossRef]
34. X. Ye and M. D. McCluskey, “Subtractive imaging using Gaussian fits and image moments in confocal microscopy,” Res. J. Opt. Photon. 1(1), 1000102 (2017).
35. M. K. Hu, “Visual pattern recognition by moment invariants,” IEEE Trans. Inform. Theory 8(2), 179–187 (1962). [CrossRef]
36. J. Liang, R. N. Kohn, M. F. Becker, and D. J. Heinzen, “1.5% root-mean-square flat-intensity laser beam formed using a binary-amplitude spatial light modulator,” Appl. Opt. 48(10), 1955–1962 (2009). [CrossRef]
37. B. Mercier, J. P. Rousseau, A. Jullien, and L. Antonucci, “Nonlinear beam shaper for femtosecond laser pulses, from Gaussian to flat-top profile,” Opt. Commun. 283(14), 2900–2907 (2010). [CrossRef]
38. Y. Ren, R. Lu, and L. Gong, “Tailoring light with a digital micromirror device,” Ann. Phys. 527(7-8), 447–470 (2015). [CrossRef]
39. J. B. Pawley, “Points, pixels, and gray levels: digitizing image data,” in Handbook of Biological Confocal Microscopy, J. B. Pawley, ed. (Springer, 2006).
40. S. Preibisch, S. Saalfeld, and P. Tomancak, “Globally optimal stitching of tiled 3D microscopic image acquisitions,” Bioinformatics 25(11), 1463–1465 (2009). [CrossRef]
41. M. Sivaguru, G. Fried, B. S. Sivaguru, V. A. Sivaguru, X. Lu, K. H. Choi, M. T. A. Saif, B. Lin, and S. Sadayappan, “Cardiac muscle organization revealed in 3-d by imaging whole-mount mouse hearts using two photon fluorescence and confocal microscopy,” Bio. Techniques 59(5), 295–308 (2015). [CrossRef]
42. C. M. Chia, H. C. Wang, J. A. Yeh, D. Bhattacharya, and Y. Luo, “Multiplexed holographic non-axial-scanning slit confocal fluorescence microscopy,” Opt. Express 26(11), 14288–14294 (2018). [CrossRef]
43. Y.H. Chia, J. A. Yeh, Y. Y. Huang, and Y. Luo, “Simultaneous multi-color optical sectioning fluorescence microscopy with wavelength-coded volume holographic gratings,” Opt. Express 28(25), 37177–37187 (2020). [CrossRef]
