This PDF file contains the front matter associated with SPIE Proceedings Volume 9713, including the Title Page, Copyright information, Table of Contents, and Conference Committee listing.

There are basically two types of microscope, which we call conventional and scanning. The former type is a full-field imaging system. In the latter type, the object is illuminated with a probe beam, and a signal detected. We can generalize the probe to a patterned illumination. Similarly we can generalize the detection to a patterned detection. Combining these we get a range of different modalities: confocal microscopy, structured illumination (with full-field imaging), spinning disk (with multiple illumination points), and so on. The combination allows the spatial frequency bandwidth of the system to be doubled. In general we can record a four dimensional (4D) image of a 2D object (or a 6D image from a 3D object, using an acoustic tuneable lens). The optimum way to directly reconstruct the resulting image is by image scanning microscopy (ISM). But the 4D image is highly redundant, so deconvolution-based approaches are also relevant.

ISM can be performed in fluorescence, bright field or interference microscopy. Several different implementations have been described, with associated advantages and disadvantages. In two-photon microscopy, the illumination and detection point spread functions are very different. This is also the case when using pupil filters or when there is a large Stokes shift.

We present a novel super-resolution spectral encoding of spatial frequency (srSESF) approach to break the diffraction limit and dramatically improve resolution in lateral direction. The idea is to utilize additional information about the internal structure of the object to resolve features in the lateral direction. The novel contrast mechanism is realized by reconstruction and comparison of the axial spatial frequency (period) profiles at each image point to form super-resolution image. As a result, small features, unresolved by conventional microscopy, can be visualized. Numerical simulation and experiments confirm the super-resolution abilities of the srSESF approach without applying any labels.

MultiPhoton SPatIal Frequency modulated Imaging (MP-SPIFI) has recently demonstrated the ability to simultaneously obtain super-resolved images in both coherent and incoherent scattering processes — namely, second harmonic generation and two-photon fluorescence, respectively.¹ In our previous analysis, we considered image formation produced by the zero and first diffracted orders from the SPIFI modulator. However, the modulator is a binary amplitude mask, and therefore produces multiple diffracted orders. In this work, we extend our analysis to image formation in the presence of higher diffracted orders. We find that tuning the mask duty cycle offers a measure of control over the shape of super-resolved point spread functions in an MP-SPIFI microscope.

Surpassing the resolution of optical microscopy defined by the Abbe diffraction limit, while simultaneously achieving optical sectioning, is a challenging problem particularly for live cell imaging of thick samples. Among a few developing techniques, structured illumination microscopy (SIM) addresses this challenge by imposing higher frequency information into the observable frequency band confined by the optical transfer function (OTF) of a conventional microscope either doubling the spatial resolution or filling the missing cone based on the spatial frequency of the pattern when the patterned illumination is two-dimensional. Standard reconstruction methods for SIM decompose the low and high frequency components from the recorded low-resolution images and then combine them to reach a high-resolution image. In contrast, model-based approaches rely on iterative optimization approaches to minimize the error between estimated and forward images. In this paper, we study the performance of both groups of methods by simulating fluorescence microscopy images from different type of objects (ranging from simulated two-point sources to extended objects). These simulations are used to investigate the methods' effectiveness on restoring objects with various types of power spectrum when modulation frequency of the patterned illumination is changing from zero to the incoherent cut-off frequency of the imaging system. Our results show that increasing the amount of imposed information by using a higher modulation frequency of the illumination pattern does not always yield a better restoration performance, which was found to be depended on the underlying object. Results from model-based restoration show performance improvement, quantified by an up to 62% drop in the mean square error compared to standard reconstruction, with increasing modulation frequency. However, we found cases for which results obtained with standard reconstruction methods do not always follow the same trend.

Random structured illumination patterns are used to demonstrate effective sectioning as well as super-resolution images in conjunction with an incoherent light source. By projecting patterns of varied spatial frequencies and using blind deconvolution of an unknown point spread function, super-resolution is achieved. Random patterns produce more consistent sectioning and super-resolution given an unknown optical transfer function. Further, using a randomly distributed pattern provides a low cost solution to obtaining information similar to that produced in confocal microscopy and other methods of structured illumination, without the requirement of precise projection patterns, coherent light sources, or fluorescence.

Three-dimensional real-time imaging is a fundamental challenge for understanding dynamic processes. In this presentation, we propose a new method based on a tunable acoustic gradient lens integrated in a simple optical system. By synchronizing a pulsed LED with a high-speed camera, we are able to resolve a volume of 2 millimeters and 2 millimeters with depth 1 millimeter in 7 microseconds. A simulation model of the optical system is provided and serves as a useful tool for designing the optical system for the desired aspect-ratio of the imaging volume. The ability to resolve a volume in microseconds opens the door to exploring the fundamental dynamics in micro scale.

Focal Modulation microscopy (FMM) is a novel imaging technique offering enhanced optical sectioning. FMM introduces spatiotemporal phase modulation in the illumination beam, resulting in intensity modulated excitation and emission light. As the background fluorescence excited by scattered photons are stationary, it is possible to differentiate it from the signal. Currently we are exploring a high-speed implementation of FMM. We have developed line scan focal modulated microscopy, which features parallel illumination and parallel detection. An imaging speed of 100 frames per second achieved with such a prototype. We have conducted a series of experiments to compare the performances of line-scan FMM, line-scan confocal microscopy, and point-scanning confocal microscopy. It is evident that line-scan FMM provides the best solution for a combination of high-speed, high contrast, and high spatial resolution.

Optical imaging in a turbid medium is limited because of multiple scattering a photon undergoes while traveling through the medium. Therefore, optical imaging is unable to provide high resolution information deep in the medium. In the case of soft tissue, acoustic waves unlike light, can travel through the medium with negligible scattering. However, acoustic waves cannot provide medically relevant contrast as good as light. Hybrid solutions have been applied to use the benefits of both imaging methods. A focused acoustic wave generates a force inside an acoustically absorbing medium known as acoustic radiation force (ARF). ARF induces particle displacement within the medium. The amount of displacement is a function of mechanical properties of the medium and the applied force. To monitor the displacement induced by the ARF, speckle pattern analysis can be used. The speckle pattern is the result of interfering optical waves with different phases. As light travels through the medium, it undergoes several scattering events. Hence, it generates different scattering paths which depends on the location of the particles. Light waves that travel along these paths have different phases (different optical path lengths). ARF induces displacement to scatterers within the acoustic focal volume, and changes the optical path length. In addition, temperature rise due to conversion of absorbed acoustic energy to heat, changes the index of refraction and therefore, changes the optical path length of the scattering paths. The result is a change in the speckle pattern. Results suggest that the average change in the speckle pattern measures the displacement of particles and temperature rise within the acoustic wave focal area, hence can provide mechanical and thermal properties of the medium.

Widefield and confocal fluorescence microscopy using a single objective suffer from poor resolution and a strong anisotropy between the lateral and axial resolution. Coherently combining the excitation and emission from two coaxial objectives improves the axial resolution up to sevenfold, but leaves the lateral resolution unchanged. Here we investigate the coherent combination of three objectives to create a point spread function (PSF) that is isotropic with higher resolution in the plane of the objectives. We develop a theoretical framework for simulating the performance of interferometric imaging with three objectives. Using three identical objectives with a large working distance and 0.9 numerical aperture (NA), the full-width half maximum of the confocal PSF is 135 nm compared to the lateral FWHM of 274 nm for imaging with a single objective at a wavelength of 515 nm.

Coherent holographic image reconstruction by phase transfer (CHIRPT) is an imaging method that permits digital holographic propagation of fluorescent light. The image formation process in CHIRPT is based on illuminating the specimen with a precisely controlled spatio-temporally varying intensity pattern. This pattern is formed by focusing a spatially coherent illumination beam to a line focus on a spinning modulation mask, and image relaying the mask plane to the focal plane of an objective lens. Deviations from the designed spatio-temporal illumination pattern due to imperfect mounting of the circular modulation mask onto the rotation motor induce aberrations in the recovered image. Here we show that these aberrations can be measured and removed non-iteratively by measuring the disk aberration phase externally. We also demonstrate measurement and correction of systematic optical aberrations in the CHIRPT microscope.

I present multimodal wide-field interferometric microscopy platform for label-free 3-D imaging of live cells during fast flow. Using holographic optical tweezers, multiple cells can be optically trapped and rapidity rotated on all axes, while acquired using an external off-axis wide-field interferometric module developed in our lab. The interferometric projections are rapidly processed into the 3-D refractive-index profile of the cells using a tomographic phase microscopy algorithms that take into consideration optical diffraction effects. The algorithms for the 3-D refractive-index reconstruction, and for calculating various morphological parameters that should serve for online sorting of cells, are efficiently implemented in a nearly real-time manner. The potential of this new high-throughput imaging technique is for label-free image analysis and sorting of cells during flow, to substitute current cell sorting devices, which are based on external labeling that eventually damages the cell sample.

Digital holographic microscopy (DHM) enables high resolution non-destructive inspection of technical surfaces and minimally-invasive label-free live cell imaging. However, the analysis of confluent cell layers represents a challenge as quantitative DHM phase images in this case do not provide sufficient information for image segmentation, determination of the cellular dry mass or calculation of the cell thickness. We present novel strategies for the analysis of confluent cell layers with quantitative DHM phase contrast utilizing a histogram based-evaluation procedure. The applicability of our approach is illustrated by quantification of drug induced cell morphology changes and it is shown that the method is capable to quantify reliable global morphology changes of confluent cell layers.

With the advancement of 3D display technology, 3D imaging of macroscopic objects has drawn much attention as they provide the contents to display. The most widely used imaging methods include a depth camera, which measures time of flight for the depth discrimination, and various structured illumination techniques. However, these existing methods have poor depth resolution, which makes imaging complicated structures a difficult task. In order to resolve this issue, we propose an imaging system based upon low-coherence interferometry and off-axis digital holographic imaging. By using light source with coherence length of 200 micro, we achieved the depth resolution of 100 micro. In order to map the macroscopic objects with this high axial resolution, we installed a pair of prisms in the reference beam path for the long-range scanning of the optical path length. Specifically, one prism was fixed in position, and the other prism was mounted on a translation stage and translated in parallel to the first prism. Due to the multiple internal reflections between the two prisms, the overall path length was elongated by a factor of 50. In this way, we could cover a depth range more than 1 meter. In addition, we employed multiple speckle illuminations and incoherent averaging of the acquired holographic images for reducing the specular reflections from the target surface. Using this newly developed system, we performed imaging targets with multiple different layers and demonstrated imaging targets hidden behind the scattering layers. The method was also applied to imaging targets located around the corner.

We propose a new method for three-dimensional (3-D) imaging without depth scanning that we refer to as the dual-detection confocal microscopy (DDCM). Compared to conventional confocal microscopy, DDCM utilizes two pinholes of different sizes. DDCM generates two axial response curves which have different stiffness according to the pinhole diameters. The two axial response curves can draw the characteristics curve of the system which shows the relationship between the axial position of the sample and the intensity ratio. Utilizing the characteristic curve, the DDCM reconstructs a 3-D surface profile with a single 2-D scanning. The height of each pixel is calculated by the intensity ratio of the pixel and the intensity ratio curve. Since the height information can be obtained directly from the characteristic curve without depth scanning, a major advantage of DDCM over the conventional confocal microscopy is a speed. The 3-D surface profiling time is dramatically reduced. Furthermore, DDCM can measure 3-D images without the influence of the sample condition since the intensity ratio is independent of the quantum yield and reflectance. We present two types of DDCM, such as a fluorescence microscopy and a reflectance microscopy. In addition, we extend the measurement range axially by varying the pupil function. Here, we demonstrate the working principle of DDCM and the feasibility of the proposed methods.

We present a volumetric imaging method for biological tissue that is free of mechanically scanning components. The optical sectioning in the system is obtained by structured illumination microscopy (SIM) with the depth of focus being varied by the use of an electronic tunable-focus lens (ETL). The performance of the axial scanning mechanism was evaluated and characterized in conjunction with SIM to ensure volumetric images could be recorded and reconstructed without significant losses in optical section thickness and lateral resolution over the full desired scan range. It was demonstrated that sub-cellular image resolutions were obtainable in both microsphere films and in ex vivo oral mucosa, spanning multiple cell layers, without significant losses in image quality. The mechanism proposed here has the ability to be integrated into any wide-field microscopy system to convert it into a three-dimensional imaging platform without the need for axial scanning of the sample or imaging optics. The ability to axially scan independent of mechanical movement also provides the opportunity for the development of endoscopic systems which can create volumetric images of tissue in vivo.

We present results of the development of a non-contacting instrument, called fScan, based on scanning confocal fluorometry for assessing the diffusion of materials through a tissue matrix.

There are many areas in healthcare diagnostics and screening where it is now widely accepted that the need for new quantitative monitoring technologies is a major pinch point in patient diagnostics and in vitro testing. With the increasing need to interpret 3D responses this commonly involves the need to track the diffusion of compounds, pharma-active species and cells through a 3D matrix of tissue. Methods are available but to support the advances that are currently only promised, this monitoring needs to be real-time, non-invasive, and economical. At the moment commercial meters tend to be invasive and usually require a sample of the medium to be removed and processed prior to testing. This methodology clearly has a number of significant disadvantages.

fScan combines a fiber based optical arrangement with a compact, free space optical front end that has been integrated so that the sample’s diffusion can be measured without interference. This architecture is particularly important due to the "wet" nature of the samples. fScan is designed to measure constructs located within standard well plates and a 2-D motion stage locates the required sample with respect to the measurement system.

Results are presented that show how the meter has been used to evaluate movements of samples through collagen constructs in situ without disturbing their kinetic characteristics. These kinetics were little understood prior to these measurements.

We present a method to acquire both fluorescence and high-resolution bright-field images with correction for the spatially varying aberrations over a microscope’s wide field-of-view (FOV). First, the procedure applies Fourier ptychographic microscopy (FPM) to retrieve the amplitude and phase of a sample, at a resolution that significantly exceeds the cutoff frequency of the microscope objective lens. At the same time, FPM algorithm is able to leverage on the redundancy within the set of acquired FPM bright-field images to estimate the microscope aberrations, which usually deteriorate in regions further away from the FOV’s center. Second, the procedure acquires a raw wide-FOV fluorescence image within the same setup. Lack of moving parts allows us to use the FPM-estimated aberration map to computationally correct for the aberrations in the fluorescence image through deconvolution. Overlaying the aberration-corrected fluorescence image on top of the high-resolution bright-field image can be done with accurate spatial correspondence. This can provide means to identifying fluorescent regions of interest within the context of the sample’s bright-field information. An experimental demonstration successfully improves the bright-field resolution of fixed, stained and fluorescently tagged HeLa cells by a factor of 4.9, and reduces the error caused by aberrations in a fluorescence image by 31%, over a field of view of 6.2 mm by 9.3 mm. For optimal deconvolution, we show the fluorescence image needs to have a signal-to-noise ratio of ~18.

The modulation of the state of polarization of photons due to scatter generates associated geometric phase that is being investigated as a means for decreasing the degree of uncertainty in back-projecting the paths traversed by photons detected in backscattered geometry. In our previous work, we established that polarimetrically detected Berry phase correlates with the mean photon penetration depth of the backscattered photons collected for image formation. In this work, we report on the impact of state-of-linear-polarization (SOLP) filtering on both the magnitude and population distributions of image forming detected photons as a function of the absorption coefficient of the scattering sample. The results, based on Berry phase tracking implemented Polarized Monte Carlo Code, indicate that sample absorption plays a significant role in the mean depth attained by the image forming backscattered detected photons.

The abstract is not available

A robust method of creation of panoramic images which does not consume much processing resources is proposed and investigated. The phase correlation (PC) method is taken as a basic one because of simplicity of its algorithm and low computing time due to application of FFT technique. Standard PC procedure is modified by preprocessing of source frames of panoramic images in spatial domain. Preprocessing comprises Linear-Scale Differential Analysis (LSDA) with sequent content-dependent thresholding of intensity gradients. Method is proved for artificially blurred and noise corrupted images. It is shown that new robust algorithm allows to increase the productivity of creation of panoramic images keeping the probability of successive stitching close to maximum even for low-quality source frames.

We report on improving lateral resolution of optical coherence tomography (OCT) for imaging of skins using multiframe superresolution technique. Through introduction of suitable slight transverse positional shifts among a series of C-scans, the superresolution processing of the lateral low resolution images at each axial depth reconstructs a high resolution image. Superresolution processing of all depth layers yields a high resolution 3D image. Using known resolution photomasks, 3 times lateral resolution improvement has been confirmed for both low and high numerical aperture OCT imaging. The superresolution processed OCT 3D skin image provides much more feature details for all subsurface depth layers within the OCT axial imaging range.

Cyclic AMP (cAMP) is a ubiquitous second messenger known to differentially regulate many cellular functions over a wide range of timescales. Several lines of evidence have suggested that the distribution of cAMP within cells is not uniform, and that cAMP compartmentalization is largely responsible for signaling specificity within the cAMP signaling pathway. However, to date, no studies have experimentally measured three dimensional (3D) cAMP distributions within cells. Here we use both 2D and 3D hyperspectral microscopy to visualize cAMP gradients in endothelial cells from the pulmonary microvasculature (PMVECs). cAMP levels were measured using a FRETbased cAMP sensor comprised of a cAMP binding domain from EPAC sandwiched between FRET donors and acceptors — Turquoise and Venus fluorescent proteins. Data were acquired using either a Nikon A1R spectral confocal microscope or custom spectral microscopy system. Analysis of hyperspectral image stacks from a single confocal slice or from summed images of all slices (2D analysis) indicated little or no cAMP gradients were formed within PMVECs under basal conditions or following agonist treatment. However, analysis of hyperspectral image stacks from 3D cellular geometries (z stacks) demonstrate marked cAMP gradients from the apical to basolateral membrane of PMVECs. These results strongly suggest that 2D imaging studies of cAMP compartmentalization — whether epifluorescence or confocal microscopy — may lead to erroneous conclusions about the existence of cAMP gradients, and that 3D studies are required to assess mechanisms of signaling specificity.

We here show 3D light sheet microscopy images of fixed and cleared murine colon tissue in-toto, which offer relevant cellular information without the need for physically sectioning the tissue. We have applied the recently developed CUBIC protocol (Susaki et al. Cell 157:726, 2014) for colon tissues and have found that this clearing protocol enables imaging all the way to the central part of the lumen with cellular resolution, thus opening new ways for 3D imaging of colon samples.

Here we first propose a fast, one-shot, non-Bayesian method which performs a numerical synthesis of a moving aperture in order to reduce the noise in Digital Holography without prior information on its statistics. Starting from one single hologram capture, multiple uncorrelated reconstructions are provided by random sparse resampling masks, which can be incoherently averaged. Thus, the problem of the setup complexity introduced by multiple recordings gets solved. Besides, at the scope of performing DH display using a SLM, it is highly required to operate directly on the hologram, in order to obtain its denoised version without losing the coherence between amplitude and phase information. We then move a step forward, showing a novel encoding formula allowing us to directly synthesize denoised holograms to be optically displayed by SLMs.

Microscope lenses can have either large field of view (FOV) or high resolution, not both. Fourier ptychographic microscopy (FPM) is a new computational microscopy technique that circumvents this limit by fusing information from multiple images taken with different illumination angles. The result is a Gigapixel-scale image having both wide FOV and high resolution, i.e. large space-bandwidth product. FPM has enormous potential for revolutionizing microscopy and has already found application in digital pathology. However, it suffers from long acquisition times (on the order of minutes), limiting throughput. Faster capture times would not only improve imaging speed, but also allow studies of live samples, where motion artifacts degrade results. Here, we present a new source coding scheme to improve the acquisition time by several orders of magnitude, enabling high space-bandwidth-time product imaging. We demonstrate our high-speed Gigapixel phase microscopy method by imaging both growing and confluent in vitro cell cultures, capturing videos of subcellular dynamical phenomena in popular cell lines undergoing division and migration. Further, we extend the Gigapixel imaging capability to 3D by processing 4D light field measurements from sequential illumination scanning. Starting from geometric optics light field refocusing, we incorporate phase retrieval and correct diffraction artifacts using a multislice coherent model that accounts for multiple scattering. Further, we incorporate dark-field images to achieve lateral resolution beyond the diffraction limit of the objective (5× larger NA) and optical sectioning better than the depth of field, using a low-magnification objective with a large FOV.

We demonstrate that super-resolution imaging of specimens containing sub-diffraction-limited features is feasible by using dielectric microwires fabricated through capillary force lithography followed by photopatterning. As supplementary micron scale cylindrical lenses, we fabricated uniform-sized microwires with and 5 and 10 μm diameters and refractive index ~1.3-1.6. The microwires are placed in contact with the specimen to collect the information of the sub-wavelength features of the specimen and transmit them to the far-field with magnification enabling imaging with two-fold resolution improvement. Potential applications of our imaging technique include biological imaging, microfluidics, and nanophotonics applications.

A high-magnification image of a biological sample can generally be obtained by an optical microscope with an objective lens, moving the image sensor with a sub-pixel shift and the subsequent image processing for super-resolution. However, to obtain a high-resolution image, a large number of images will be required for the super-resolution, and thus it is difficult to achieve real-time operation, and the field-of-view (FOV) is not sufficiently wide. The currently proposed digital holography technique places a sample on the image sensor and captures the interference fringe (hologram) to reconstruct a 3D high-resolution image in a computer. This technique ensures the features of a wide FOV, whereas the high resolution obtained by image processing cannot ensure real-time operation, because it requires recursive calculations of light propagation and adequate computer resources. To realize wide FOV and the real-time operation at the same time, we have developed a new technique: Lensfree on-chip high-resolution imaging using two-way lighting. High-resolution image is immediately obtained by image processing of the low-resolution images of the samples. This makes it possible to ensure a wide FOV, a deep depth of focus without the need for focus adjustment, and a continuously expanding operation. We also discuss the limitations of the high resolution.

We present a wide-field fluorescence lifetime imaging (FLIM) system with optical sectioning by structured illumination microscopy (SIM). FLIM measurements were made using a time gated ICCD camera in conjunction with a pulsed nitrogen dye laser operating at 450 nm. Intensity images were acquired at multiple time delays from a trigger initiated by a laser pulse to create a wide-field FLIM image, which was then combined with three phase SIM to provide optical sectioning. Such a mechanism has the potential to increase the reliability and accuracy of the FLIM measurements by rejecting background intensity. SIM also provides the opportunity to create volumetric FLIM images with the incorporation of scanning mechanisms for the sample plane. We present multiple embodiments of such a system: one as a free space endoscope and the other as a fiber microendoscope enabled by the introduction of a fiber bundle. Finally, we demonstrate the efficacy of such an imaging system by imaging dyes embedded in a tissue phantom.

This work presents a technique for a full 3D imaging of biological samples tagged with gold-nanoparticles (GNPs) using only two images, rather than many images per volume as is currently needed for 3D optical sectioning microscopy. The proposed approach is based on the Gerchberg-Saxton (GS) phase retrieval algorithm. The reconstructed field is free space propagated to all other focus planes using post processing, and the 2D z-stack is merged to create a 3D image of the sample with high fidelity. Because we propose to apply the phase retrieving on nano particles, the regular ambiguities typical to the Gerchberg-Saxton algorithm, are eliminated. In addition, since the method requires the capturing of two images only, it can be suitable for 3D live cell imaging. The proposed concept is presented and validated both on simulated data as well as experimentally.

Compressive Sensing (CS)-based technologies have shown potential to improve the efficiency of acquisition, manipulation, analysis and storage processes on signals and imagery with slight discernible loss in data performance. The CS framework relies on the reconstruction of signals that are presumed sparse in some domain, from a significantly small data collection of linear projections of the signal of interest. As a result, a solution to the underdetermined linear system resulting from this paradigm makes it possible to estimate the original signal with high accuracy. One common approach to solve the linear system is based on methods that minimize the L1-norm. Several fast algorithms have been developed for this purpose. This paper presents a study on the use of CS in high-resolution reflectance confocal microscopy (RCM) images of the skin. RCM offers a cell resolution level similar to that used in histology to identify cellular patterns for diagnosis of skin diseases. However, imaging of large areas (required for effective clinical evaluation) at such high-resolution can turn image capturing, processing and storage processes into a time consuming procedure, which may pose a limitation for use in clinical settings. We present an analysis on the compression ratio that may allow for a simpler capturing approach while reconstructing the required cellular resolution for clinical use. We provide a comparative study in compressive sensing and estimate its effectiveness in terms of compression ratio vs. image reconstruction accuracy. Preliminary results show that by using as little as 25% of the original number of samples, cellular resolution may be reconstructed with high accuracy.

In super-resolution microscopy, we use fluorescence depletion, where an erase beam quenches a molecule in the S₁ state generated by a pump beam, and then prevents fluorescence from the S₁ state. When a tight doughnut shaped erase beam with is focused on the dyed sample together with a Gaussian pump beam, the remaining fluorescence spot in the focal plane becomes smaller than the diffraction-limited size. Applying destructive interference to the erase beam, erase beam has a minute three-dimensional dark spot surrounded by the light near the focal region. Since this spot introduces fluorescence depletion along the optical axis as in the focal plane, we can achieve three-dimensional super-resolution microscopy. However, to overcome the diffraction limit, an extremely precise optical alignment is required for projecting the focused pump beam into the dark spot of the erase beam. To resolve this technical issue, we fabricated a two-color annular hybrid wave plate (TAHWP) by combining two multi-order wave quartz plates. Although the pump and erase beams co-axially pass through the plate; the pump beam retains its original Gaussian shape, while the erase beam undergoes destructive interference. Inserting the TAHWP into a commercial scanning laser microscope, a three-dimensional spherical fluorescence spot with a volume of (~100 nm)³ can be created. Beside eliminating alignment problems and yielding a compact setup, the TAHWP makes our proposed method very suitable for commercial microscope systems. In this study, we report about detailed fabrication procedure and three-dimensional image properties given by the TAHWP.

A beam-scanning microscope is described based on a temporally multiplexed Lissajous trajectory for achieving 1 kHz frame rate 3D imaging. The microscope utilizes two fast-scan resonant mirrors to direct the optical beam on a circuitous, Lissajous trajectory through the field of view. Acquisition of two simultaneous focal planes is achieved by implementation of an optical delay line, producing a second incident beam at a different focal plane relative to the initial incident beam. High frame rates are achieved by separating the full time-domain data into shorter sub-trajectories resulting in undersampling of the field of view. A model-based image reconstruction (MBIR) 3D in-painting algorithm is utilized for interpolating the missing data to recover full images. The MBIR algorithm uses a maximum a posteriori estimation with a generalized Gaussian Markov random field prior model for image interpolation. Because images are acquired using photomultiplier tubes or photodiodes, parallelization for multi-channel imaging is straightforward.

Preliminary results obtained using a Lissajous trajectory beam-scanning approach coupled with temporal multiplexing by the implementation of an optical delay line demonstrate the ability to acquire 2 distinct focal planes simultaneously at frame rates >450 Hz for full 512 × 512 images. The use of multi-channel data acquisition cards allows for simultaneous multimodal image acquisition with perfect image registry between all imaging modalities. Also discussed here is the implementation of Lissajous trajectory beam-scanning on commercially available microscope hardware.

We present a deformable mirror based remote focusing method for three-dimensional imaging in high-resolution microscopy systems. The method relies on predefined mirror mode arrays that are obtained during initial mirror training step with a low complexity wavefront-sensing module. The imaging plane can be refocused over distances over a hundred times greater than the original depth of field of the objective lens along the optical axis at millisecond rates. We will demonstrate the combination of the remote focusing method with spatiotemporally focused two-photon excitation applied to three-dimensional imaging of biological samples.

We present a compound adaptive objective lens in which a water-filled membrane lens is inserted into a front group (one lens) and a back group (two lenses). This adaptive objective lens works in the ultrabroad near infrared waveband (760nm ~ 920nm) with the volume scan of > 1mm³ and the resolution of 2.8 μm (calculated at the wavelength of 840 nm). The focal range is 19.5mm ~ 20.5mm and the numerical number is 0.196. The size of the adaptive lens is 10mm (diameter) × 17mm (length). This kind of lens can be widely used in three-dimensional (3D) volume biomedical imaging instruments, such as confocal microscope, optical coherence tomography (OCT), two photon microscope, etc.

We present a simple technique which uses a random phase object for single-shot characterization of an optical system's phase transfer function. Existing methods for aberration measurement typically involve holography, requiring complicated wavefront sensing optics or through-focus measurements with known test objects (e.g. pinholes, fluorescent beads) for pupil recovery from the measured wavefront. Here, it is demonstrated that a weak diffuser can be used to recover the pupil of an imaging system in a single measurement, without exact knowledge of the diffuser's surface. Due to its stochastic nature, the diffuser scatters light to a wide range of spatial frequencies, thus probing the entire pupil plane. A linear theory based on the weak object approximations predicts the spectrum of the measured speckle intensity to depend directly on the pupil function. Numerical simulations of diffusers with varying strength confirm the validity of the theory and indicate sufficient conditions under which diffusers act as weak phase objects. Using index matching oils to modulate diffuser strength, experiments are shown to successfully recover aberrations from an optical system using coherent illumination. Additionally, this technique is applied to the recovery of defocus in images of a weak phase object obtained through a commercial microscope under partially coherent illumination.

In confocal microscopy the polarization of the illumination beam plays an important role in determining the orientation of the fluorescent molecules being illuminated. The efficiency of the excitation depends on the angle between the excitation electric field and the direction of the molecular dipole. In order to determine the orientation of the fluorescent molecules in the focal plane the molecules are to be excited using two mutually orthogonal electric fields. In this paper we show how a computer generated holography technique can be implemented using a ferroelectric liquid crystal spatial light modulator to conveniently obtain two images of the same target once with an X polarized illumination beam and another with a Y polarized illumination beam.

There is significant interest in the scientific community to develop a reliable and precise quantification of the direction of collagen fibers within the cornea in order to learn more about how collagen contributes to the cornea's shape and structure. Previous work has shown that quantification of these fibers' orientations is possible using Second Harmonic Generation (SHG) microscopy, a modality that utilizes the non-centrosymmetric properties of collagen to obtain details of their macromolecular structure. Many attempts at using SHG to this purpose result in whole volume approximations which do not take into account variations through the depth of the cornea. Additionally, other algorithms have used non-linear processes which limit computation time and result in sensitivity to sample area, resolution, and illumination. We propose a linear method for quantification of collagen fiber orientation which utilizes the precise sectioning properties of SHG and is independent of non-linear artifacts to aid in further classification of cornea structure.

Oblique plane microscopy (OPM) is a light-sheet fluorescence microscopy technique that is implemented on a standard inverted microscope frame. OPM uses a single high numerical aperture microscope objective to both produce a tilted excitation light-sheet and to image the fluorescence emitted from the tilted plane back to the cameras. It is therefore compatible with conventional sample-mounting techniques such as microscope slides and multiwell plates. Four excitation laser lines and two high-speed sCMOS cameras with separate emission filters enable the simultaneous imaging of several fluorophores and spectral ratiometric FRET acquisitions. Previously, 3-D OPM imaging has been implemented by remote refocusing. Here, a stage-scanning approach to 3-D OPM imaging is demonstrated - enabling three-dimensional multi-channel acquisition including of multiwell plates - and the synchronization of the stage movement and camera acquisition will be described. The ability of the stage-scanning system to image fields of view larger than the field of view of the primary microscope objective is demonstrated using fluorescently labelled limbs of crustaceans and its ability to perform time-lapse 3-D imaging over 12 hours is demonstrated using a sample of tumor spheroids with an acquisition time of 3 s for a typical spheroid providing 400x1280x1024 voxels per spheroid. We also apply the system to spectral ratiometric Förster resonant energy transfer (FRET) measurements in tumor spheroids expressing a FRET biosensor and in a 96-well plate seeded with cell samples expressing varying concentrations of a FRETting and non-FRETting constructs.

Imaging depth is one of the most prominent limitations in light microscopy. The depth in which we are still able to resolve biological structures is limited by the scattering of light within the sample. We have developed an algorithm to compensate for the influence of scattering. The potential of algorithm is demonstrated on a 3D image stack of a zebrafish embryo captured with a selective plane illumination microscope (SPIM). With our algorithm we were able shift the point in depth, where scattering starts to blur the imaging and effect the image quality by around 30 µm. For the reconstruction the algorithm only uses information from within the image stack. Therefore the algorithm can be applied on the image data from every SPIM system without further hardware adaption. Also there is no need for multiple scans from different views to perform the reconstruction. The underlying model estimates the recorded image as a convolution between the distribution of fluorophores and a point spread function, which describes the blur due to scattering. Our algorithm performs a space-variant blind deconvolution on the image. To account for the increasing amount of scattering in deeper tissue, we introduce a new regularizer which models the increasing width of the point spread function in order to improve the image quality in the depth of the sample. Since the assumptions the algorithm is based on are not limited to SPIM images the algorithm should also be able to work on other imaging techniques which provide a 3D image volume.

Wavefront-engineered microscope with greatly extended depth of field (EDoF) is designed and demonstrated for volumetric imaging with near-diffraction limited optical performance. A bright field infinity-corrected transmissive/reflective light microscope is built with Kohler illumination. A home-made phase mask is placed in between the objective lens and the tube lens for ease of use. General polynomial function is adopted in the design of the phase plate for robustness and custom merit function is used in Zemax for optimization. The resulting EDoF system achieves an engineered point spread function (PSF) that is much less sensitive to object depth variation than conventional systems and therefore 3D volumetric information can be acquired in a single frame with expanded tolerance of defocus. In Zemax simulation for a setup using 32X objective (NA = 0.6), the EDoF is 20μm whereas a conventional one has a DoF of 1.5μm, indicating a 13 times increase. In experiment, a 20X objective lens with NA = 0.4 was used and the corresponding phase plate was designed and fabricated. Retinal fluorescence images of the EDoF microscope using passive adaptive optical phase element illustrate a DoF around 100μm and it is able to recover the volumetric fluorescence images that are almost identical to in-focus images after post processing. The image obtained from the EDoF microscope is also better in resolution and contrast, and the retinal structure is better defined. Hence, due to its high tolerance of defocus and fine restored image quality, EDoF optical systems have promising potential in consumer portable medical imaging devices where user’s ability to achieve focus is not optimal, and other medical imaging equipment where achieving best focus is not a necessary.

We present our latest advances in highly dynamic photothermal interferometric phase microscopy for quantitative, selective contrast imaging. Gold nanoparticles can be bio-functionalized to bind specific cells. When stimulating gold nanoparticles at their plasmon-peak wavelength, local increase of temperature occurs due to plasmon resonance. This causes a rapid change of optical phase of the light beam interacting with the sample. These phase changes can be recorded by interferometric phase microscopy and analyzed to form a photothermal image of the binding sites of the nanoparticles in the cells. Furthermore, by increasing the excitation laser light, one can deplete certain cells at will. Usually, the analysis of the photothermal signals utilizes a Fourier transform, which is computational time consuming. This makes photothermal imaging not suitable for applications requiring dynamic imaging or real-time quantitative analysis, such as for analyzing and sorting cells during their fast flow. For this goal, we have developed new algorithms, based on discrete Fourier transform variants, enabling fast analysis of photothermal signals from nanoparticles in live and highly dynamic cells. For the first time, video-rate photothermal signals are obtained, which forms the basis for real-time interferometric phase microscopy with molecular specificity. This technique holds great potential for using photothermal imaging in flow cytometry.

Holographic phase microscopy has seen rapid growth in the past two decades. Numerous schemes have been proposed and commercial products are now available. Since most systems are laser based, speckle noise and other non-signal interference in the system have been problematic, limiting the technique’s phase sensitivity, image quality and the ability for accurate quantitative analysis. Low coherence source-based HPM have also been proposed to mitigate this issue, but often with increased system complexity and reduced implementation flexibility. Here, we demonstrate a swept-source HPM technique, which acquires on-axis holograms while continuously scanning the laser through a range of wavelengths. This technique is capable of identifying interference from various sources and effectively isolating sample interference, therefore minimizing unwanted signals and achieving high spatial and temporal sensitivity across the entire field of view. The ability of acquiring spectral interferogram for each pixel also make it possible to implement spectral shaping, which can further suppress interference side-lobes and improve sensitivity. Additionally, when coupled with a spectral modulation technique, such interference spectrum will permit spectroscopic measurement of phase-related properties of the sample. We will introduce the principle of the system, discuss its theoretical sensitivity bound, and present its application to phase imaging of live cells.

A spectral multiplexing interferometry (SXI) method is presented for integrated birefringence and phase gradient measurement on label-free biological specimens. With SXI, the retardation and orientation of sample birefringence are simultaneously encoded onto two separate spectral carrier waves, generated by a crystal retarder oriented at a specific angle. Thus sufficient information for birefringence determination can be obtained from a single interference spectrum, eliminating the need for multiple acquisitions with mechanical rotation or electrical modulation. In addition, with the insertion of a Nomarski prism, the setup can then acquire quantitative differential interference contrast images. Red blood cells infected by malaria parasites are imaged for birefringence retardation as well as phase gradient. The results demonstrate that the SXI approach can achieve both quantitative phase imaging and birefringence imaging with a single, high-sensitivity system.

In the biological sciences, there is much emphasis on elucidating the functions of various biological components and processes. To do so, advances in general microscopy have yielded various imaging modalities to probe such processes under specific visualization and contrast requirements. Examples of modalities that are popularly integrated into conventional biological studies include fluorescent, dark-field, phase-contrast, and polarization-sensitive microscopies, with each modality offering unique insights into the biological function of the sample. Often times, however, a comprehensive understanding of biological phenomena requires the integration of the unique and separate visualizations of various modalities. Unfortunately, conventional microscopes typically support only one modality and rarely allow multiple modalities to be used in conjunction. Though high-end microscopes may support multimodal visualization, they often require either mechanical (and often manual) toggling, which obstruct real-time multimodal imaging, or simultaneous detection via multiple cameras, which dramatically increases the microscope’s cost. Here, we present a one-shot technique that allows multiple imaging channels, of potentially different modalities, to be simultaneously detected by a single camera. We experimentally demonstrate this method on transparent cells that have been tagged for F-actin and nuclear fluorescence. Our multimodal system consists of 2-channel fluorescence and 1-channel quantitative-phase (QP) imaging, and clearly demonstrates ability for simultaneous fluorescent and QP visualization. Though we experimentally verify our framework using dual fluorescent/QP imaging, we emphasize that our framework for single-shot, simultaneous single-camera detection is applicable to an arbitrary number of widefield imaging modalities so long as they fulfill criteria for Fourier spectra separation, SNR, and detector dynamic range

This paper describes a new and novel phase shifting technique for qualitative as well as quantitative measurement in microscopy. We have developed a phase shifting device which is robust, inexpensive and involves no mechanical movement. In this method, phase shifting is implemented using LED array, beam splitters and defocused projection of Ronchi grating. The light from the LEDs are made incident on the beam splitters at spatially different locations. Due to variation in the geometrical distances of LEDs from the Ronchi grating and by sequentially illuminating the grating by switching on one LED at a time the phase shifted grating patterns are generated. The phase shifted structured patterns are projected onto the sample using microscopic objective lens. The phase shifted deformed patterns are recorded by a CCD camera. The initial alignment of the setup involves a simple procedure for the calibration for equal fringe width and intensity such that the phase shifted fringes are at equal phase difference. Three frame phase shifting algorithm is employed for the reconstruction of the phase map. The method described here is fully automated so that the phase shifted images are recorded just by switching of LEDs and has been used for the shape measurement of microscopic industrial objects. The analysis of the phase shifted images provides qualitative as well as quantitative information about the sample. Thus, the method is simple, robust and low cost compared to PZT devices commonly employed for phase shifting.

We describe the use of spatially incoherent illumination combined with quantitative phase imaging (QPI) [1] to make tridimensional reconstruction of semi-transparent biological samples. Quantitative phase imaging is commonly used with coherent illumination for the relatively simple interpretation of the phase measurement. We propose to use spatially incoherent illumination which is known to increase lateral and axial resolution compared to classical coherent illumination. The goal is to image thick samples with intracellular resolution [2]. The 3D volume is imaged by axially scanning the sample with a quadri-wave lateral shearing interferometer used as a conventional camera while using spatially incoherent white-light illumination (native microscope halogen source) or NIR light. We use a non-modified inverted microscope equipped with a Z-axis piezo stage. A z-stack is recorded by objective translation along the optical axis. The main advantages of this approach are its easy implementation, compared to the other state-of-the-art diffraction tomographic setups, and its speed which makes even label-free 3D living sample imaging possible. A deconvolution algorithm is used to compensate for the loss in contrast due to spatially incoherent illumination. This makes the tomographic volume phase values quantitative. Hence refractive index could be recovered from the optical slices. We will present tomographic reconstruction of cells, thick fixed tissue of few tens of micrometers using white light, and the use of NIR light to reach deeper planes in the tissue.

Using a point spread function (PSF) to localize a point-like object, such as a fluorescent molecule or microsphere, represents a common task in single molecule microscopy image data analysis. The localization may differ in purpose depending on the application or experiment, but a unifying theme is the importance of being able to closely recover the true location of the point-like object with high accuracy. We present two simulation studies, both relating to the performance of object localization via the maximum likelihood fitting of a PSF to the object's image. In the first study, we investigate the integration of the PSF over an image pixel, which represents a critical part of the localization algorithm. Specifically, we explore how the fineness of the integration affects how well a point source can be localized, and find the use of too coarse a step size to produce location estimates that are far from the true location, especially when the images are acquired at relatively low magnifications. We also propose a method for selecting an appropriate step size. In the second study, we investigate the suitability of the common practice of using a PSF to localize a microsphere, despite the mismatch between the microsphere’s image and the fitted PSF. Using criteria based on the standard errors of the mean and variance, we find the method suitable for microspheres up to 1 μm and 100 nm in diameter, when the localization is performed, respectively, with and without the simultaneous estimation of the width of the PSF.

In recent years, the techniques of super-resolution have generated widespread impacts in science. Stimulated emission depletion (STED) microscopy is known for achieving sub-diffraction-limit resolution by using a donut-shaped beam to deplete the fluorescence around a focal spot while leaving a central part active to emit fluorescence. However, since STED microscopy is based on fluorescence, it suffers from photo-bleaching. We recently developed a new technique and termed it as suppression of scattering imaging (SUSI) microscopy. It uses a STED-like setup and achieves super resolution imaging by utilizing the nonlinearity of scattering from gold nanoparticles. Therefore, SUSI microscopy avoids the photo-bleaching issue. Nonetheless, for fast volumetric imaging, SUSI microscopy is limited with slow axial translation of the objective or sample. Here we combine SUSI microscopy with a refractive-index-variable lens to axially move the focus at very high speed. This combination allows simultaneous observation of tissue dynamics over a three-dimensional volume within one second. The new technique paves the way toward high-speed super-resolution imaging for biological tissues.

One appealing aspect that compressive sensing offers is the possibility of retrieving a signal’s spectral information using a bucket detector and a characterized measurement matrix. Demonstrations of CS applied to optical coherence tomography (OCT) were performed, however, in the final signal-processing instead of the acquisition end. Here we propose a novel OCT system with a broadband superluminescent excitation and a bucket photodetector where the interferogram is obtained by spectral reconstruction. In particular, this system assumes the same interferometric setup as typical swept-source OCT systems except the excitation is replaced by a broadband source. The interferogram then passes through an off-the-shelf, fast tunable Fabry-Perot filter (FPF) of modest finesse whose free spectral range is designed to be much less than the excitation bandwidth. The spectral response is characterized a priori, before the filtered output is integrated by the photodetector. The spectral sampling measurement is repeated by altering the FPF’s resonant conditions multiple times through the cavity length. Having acquired the integrated photodetector values and the corresponding spectral filter functions, we reconstruct the original interferogram whose Fourier transform generates the tomogram. The sensitivity of this OCT technique is evaluated and compared using simulations with synthetic data. Moreover, B-scan reconstruction of the interferogram due to a fingertip was simulated using our scheme and the resultant image shows excellent reconstruction fidelity compared to the original OCT B-scan. These illustrations point towards a promising future of a new class of tomographic system which combines the respective strengths of swept-source and spectral-domain OCT.

Light induced fluorescent microscopy has long been developed to observe and understand the object at microscale, such as cellular sample. However, the transfer function of lense-based imaging system limits the resolution so that the fine and detailed structure of sample cannot be identified clearly. The techniques of resolution enhancement are fascinated to break the limit of resolution for objective given. In the past decades, the resolution enhancement imaging has been investigated through variety of strategies, including photoactivated localization microscopy (PALM), stochastic optical reconstruction microscopy (STORM), stimulated emission depletion (STED), and structure illuminated microscopy (SIM). In those methods, only SIM can intrinsically improve the resolution limit for a system without taking the structure properties of object into account. In this paper, we develop a SIM associated with Bayesian estimation, furthermore, with optical sectioning capability rendered from HiLo processing, resulting the high resolution through 3D volume. This 3D SIM can provide the optical sectioning and resolution enhancement performance, and be robust to noise owing to the Data driven Bayesian estimation reconstruction proposed. For validating the 3D SIM, we show our simulation result of algorithm, and the experimental result demonstrating the 3D resolution enhancement.

Partial differential equation (PDE)-based nonlinear diffusion processes have been widely used for image denoising. In the traditional nonlinear anisotropic diffusion denoising techniques, behavior of the diffusion depends highly on the gradient of image. However, it is difficult to get a good effect if we use these methods to reduce noise in optical coherence tomography images. Because background has the gradient that is very similar to regions of interest, so background noise will be mistaken for edge information and cannot be reduced. Therefore, nonlinear complex diffusion approaches using texture feature(NCDTF) for noise reduction in phase-resolved optical coherence tomography is proposed here, which uses texture feature in OCT images and structural OCT images to remove noise in phase-resolved OCT. Taking into account the fact that texture between background and signal region is different, which can be linked with diffusion coefficient of nonlinear complex diffusion model, we use NCDTF method to reduce noises of structure and phase images first. Then, we utilize OCT structure images to filter phase image in OCT. Finally, to validate our method, parameters such as image SNR, contrast-to-noise ratio (CNR), equivalent number of looks (ENL), and edge preservation were compared between our approach and median filter, Gaussian filter, wavelet filter, nonlinear complex diffusion filter (NCDF). Preliminary results demonstrate that NCDTF method is more effective than others in keeping edges and denoising for phase-resolved OCT.

The ability to track single fluorescent particles within a three dimensional (3D) cellular environment can provide valuable insights into cellular processes. In this paper, we present a modified nonlinear image decomposition technique called K-factor that reshapes the 3D point spread function (PSF) of an XYZ image stack into a narrow Gaussian profile. The method increases localization accuracy by ~60% with compare to regular Gaussian fitting, and improves minimal resolvable distance between overlapping PSFs by ~50%. The algorithm was tested both on simulated data and experimentally.

This work presets a novel use of the nonlinear image decomposition technique called K-factor that reshapes the three dimensional (3D) point spread function (PSF) of an XYZ image stack into a narrow Gaussian profile. The experimentally obtained PSF of a Z-stack raw data that is acquired by a widefield microscope has a more elaborate shape that is given by the Gibson and Lanni model. This shape increases the computational complexity associated with the localization routine, when used in localization microscopy techniques. Furthermore, due to its nature, this PSF spreads over a larger volume, making the problem of overlapping emitters detection more pronounced. The ability to use Gaussian fitting with high accuracy on 3D data can facilitate the computational complexity, hence reduce the processing time required for the generation of the 3D superresolved image. In addition it allows the detection of overlapping PSFs and reduces the effects of the penetration of out of focus PSFs into in focused PSFs, therefore enables the increase in the activated fluorophore density by ~50%. The algorithm was tested both on simulated data and experimentally, where it yielded an increase in the localization accuracy by ~60% with compare to regular Gaussian fitting, and improved the minimal resolvable distance between overlapping PSFs by ~50%, making it extremely applicable to the field of 3D biomedical imaging,

A novel spectral imaging technique is introduced based on a highly dispersive imaging lens system. The chromatic aberration of the lens system is utilized to spread the spectral content of the object over a focal distance. Two three-dimensional surface reconstruction algorithms, depth from focus and depth from defocus, are applied to images captured by dispersive lens system. Using these algorithms, the spectral imager is able to relate either the location of focused image or the amount of defocus at the imaging detector to the spectral content of the object. A spectral imager with ~5 nm spectral resolution is designed based on this technique. The spectral and spatial resolutions of the introduced technique are independent and can be improved simultaneously. Simulation and experimental results are presented.