This PDF file contains the front matter associated with SPIE Proceedings Volume 11245, including the Title Page, Copyright information, Table of Contents, Author and Conference Committee lists.

We have recently developed quantitative oblique back-illumination microscopy (qOBM), which enables full-field quantitative phase imaging (QPI) of objects embedded in a thick scattering medium. This epi-mode technique makes use of multiple scattering as a source of transmissive illumination from within, allowing for rich structural detail based on optical path delay. We now produce quantitative 3D renderings of index of refraction with sub- cellular detail by computing a 3D transfer function of the entire optical system, including the multiply scattered illumination, and deconvolving it from a vertical stack of phase gradient contrast images. This approach allows truly non-invasive, label-free, tomographic quantitative reconstructions if index of refraction in thick scattering samples including thick tissue samples, at low cost and with simple operation, bringing QPI’s unparalleled access to sub-cellular structural detail to previously unavailable domains of investigation.

We analyze, and test experimentally, a method of imaging at very low light levels using spatial interference between a strong local oscillator field and a weak beam. By using Fourier phase recovery techniques familiar in classical interferometry and by correcting the inhomogeneities in the reference field, we can represent the quadrature components of a given spatial mode in the complex plane. Further, by post selection filtering (convolution and sampling), the method can distinguish photons of different orbital angular momentum with a one-count standard deviation consistent with the quantum noise floor.

We demonstrate a high resolution lens-free holographic microscopy in reflection geometry based on a pixel super resolution (SR) method. The lens-free microscopy uses a novel Michelson geometry suitable to image reflective samples with the large field of view, while the Fourier domain SR technique is applied to obtain the high resolution hologram, achieving the sub-pixel resolution of 1.2 μm in the USAF reflection target by utilizing the randomly shifted low resolution images. The proposed compact microscopy technique enables to provide high resolution amplitude and phase imaging, those are suitable for biology and semiconductor imaging applications.

Although light microscopes play a pivotal role in biological discoveries, their inability to observe the workings of the tiny structures within cells – whose size fall below the classical resolution barrier – continues to frustrate biologists and optical engineers alike. It is generally assumed that the only approach to breaking the classical resolution barrier is through manipulating the fluorophores by their excitation processes, which limits their implementation to a few specific biological applications. Here, we present a new path to achieving super-resolution imaging of three-dimensional objects without the requirements of complex optics or special fluorophores; it can provide more than 3 times resolution improvement in all three dimensions. We demonstrate it on various biological samples: Drosophila brain neurons and mouse brain dendrites.

Stimulating multiple neurons simultaneously in brain is an important strategy to study neuronal networks. We propose a novel method to generate three-dimensional stimulating patterns using a Digital Micro-mirror Device (DMD). The DMD is placed in conjugate image plane. The projection pattern on the DMD changes dynamically and synchronized with a pair of galvomirrors in the conjugate plane. The three-dimensional holography is generated by 4-D control of light field (DMD controls x, y, and galvomirrors control k_x,k_y). This novel method is able to generate three-dimensional distributed illumination patterns much faster than the previous methods using Spatial Light Modulators.

We introduce a Fourier-transform hyperspectral microscope based on an ultrastable birefringent interferometer. The microscope enables wide field acquisition with broad spectral coverage, tunable spectral resolution, high sensitivity and short acquisition time. We present two prototype implementations and an add-on for a commercial microscope. We provide examples of applications in biology and solid state physics.

We propose a new hybrid scanning/wide field excitation implementation, developed on a classical inverted wide field fluorescence microscope. Compatible from epifluorescence to TIRF excitation it provides flexible and uniform illumination over wide field of views (200x200 µm²), with uniform sectioning depth. This permits to enhance TIRF imaging and control photobleaching in classical microscopy. In single molecule localization microscopy, one can enhance the localization precision by balancing the density of activated fluorophores and the photon count, and extract information from many quantitative datas thanks to the wide uniform illumination. The required laser power to achieve efficient blinking is also considerably reduced.

Spectral Domain Optical Coherence Tomography (SD-OCT) is an effective tool for volumetric imaging of collagen fiber networks, but current processing algorithms struggle to create three-dimensional models of these networks due to limited contrast and tissue complexity. We present an automated image processing algorithm that overcomes these challenges to enable quantitative visualization of three-dimensional fiber networks from OCT volumes. Samples are processed by segmenting the tissue volume surface and dividing the sample into “processing patches” which are optimally sized and oriented to fit an arbitrary volume. Fiber orientation analysis and particle filtering are used to create orientation-encoded fiber tractography. The method is demonstrated on five ex-vivo human uterine samples which were imaged as mosaic volumes using a commercial SD-OCT system, providing the first view of the three-dimensional structure of the human uterine collagen fiber network on a centimeter scale.

We report a single-shot surface three-dimensional (3-D) imaging method that uses optical coherence as a contrast mechanism to acquire the vertical (z-direction) information of an object. The illumination of the imaging system comes from a light source with the optical coherence length similar to the depth of field (DoF) of the optical system. Holographic recording is used to retrieve the coherence visibility factor, which is then converted to z-direction information. In the experiment, we compare the imaging results of our method to conventional incoherent imaging results, showing that this contrast mechanism is able to provide additional information. We also validate our 3D imaging results by using axial scanning full-field optical coherence tomography.

In this study, we implement temporal-focusing multiphoton selective excitation (TFMPSE) to light field microscopy (LFM), illuminating only the volume of interest, thus significantly reducing the background noise and providing higher contrast and accuracy for the light field image reconstruction; furthermore, offering higher penetration depth in scattering tissue via multiphoton. 3D human-skin in situ immunofluorescence images are used to demonstrate volumetric bioimaging capability. The volume rate of the TFMPSE-LFM can achieve around 100 volumes per second

In this paper, we present a single-shot multiview imaging technique for three-dimensional objects by utilizing the nature randomness of scattering media. The uncorrelated point spread functions of different scattering regions help to de-multiplex multiple elemental images at different viewing angles by deconvolution from only a single speckle pattern. Our demonstration shows that not only stereo capture with large disparity, but also, up to 7 viewing angles of a 3D object can be reconstructed with just a single shot, if corresponding PSFs are premeasured once. The elemental images are consistent with 3D object projection and images taken by multi-shot imaging.

We present an easy-to-use multi-immersion open-top light-sheet microscope designed specifically for high-throughput imaging of a diverse set of tissues prepared with a variety of clearing protocols.

Attenuation of optical fields owing to scattering and absorption limits the penetration depth into tissue. Whilst aberration correction may be used this is difficult to implement over a large field of view in heterogeneous tissue. Recently, the novel approach of attenuation-compensation of propagation-invariant light fields has shown increase in depth penetration for light-sheet microscopy. Here we show this powerful approach may be implemented in a facile manner utilizing a graded neutral density filter circumventing the need for expensive beam shaping apparatus. A ‘gold standard’ system utilizing a spatial light modulator for beam shaping is used to benchmark our low-cost implementation.

A compact, prism-based spectrograph was designed for 2-photon light sheet microscopy based on broadband, ultrashort optical pulses that corrects for chromatic aberrations and distortion. Gaussian or Bessel beams of broadband optical pulses may be used to create narrow cylindrical, nonlinear excitation volumes from which fluorophores emit characteristic spectra. In practice, a slit aperture is often used to optically section the excitation volume before being imaged onto the camera chip. Optical dispersion of the fluorescence in the transverse direction of the imaging slit allows for hyperspectral image acquisition. Hyperspectral imaging systems may be used to simultaneously image and segment multiple fluorescent reporters in biological tissue. However, current systems have properties which are undesirable for low-light microscopy including chromatic aberrations, distortion, low optical transmission, and large footprint which consumes precious laboratory real estate. Here, we present a novel spectrograph that has sufficient optical transmission, achromaticity, and distortion correction for microscopy of fluorescent reporters spanning the visible spectrum (400 – 650 nm). Across the spatial dimension of the excitation volume, the spectrograph has uniform separation of the spectral bands while maintaining a compact size and profile. A second order deconvolution algorithm is used to spectrally deconvolve overlapping fluorophores.

The goal of this work is to incorporate Convolutional Neural Networks (CNNs) into the 3D deconvolution process without training. CNNs are well suited to the problem of 2D deconvolution, however training a CNN on 3D volumes requires excessive time and impractical amounts of training data. To circumvent these problems, we use a CNN architecture as if it were a handcrafted prior, similar to the work deep image prior. Using this method, we achieve high SSIM and PSNR metrics relative to other modern techniques for deconvolving through-focus fluorescence measurements to recover a 3D volume with no training data and minimal hyperparameter tuning.

We report progress in algorithm development for a computation-based super-resolution microscopy technique. Building upon previous results, we examine our recently implemented microscope system and construct alter- native processing algorithms. Based on numerical simulations results, we evaluate the performance of each algorithm and determine the one most suitable for our super-resolution microscope.

We report a digital image refocusing framework in fluorescence microscopy (termed “Deep-Z”), where a deep neural network is trained to virtually-refocus a 2D fluorescence image onto user-defined 3D surfaces. Using Deep-Z, we demonstrated 3D reconstruction of C. elegans neuronal activity from a 2D movie, digitally increasing the depth-of-field by 20-fold. We also demonstrated digital correction of sample drift, tilt and other image aberrations, all performed after the acquisition of a single image. Deep-Z also permits cross-modality virtual refocusing, where a single 2D wide-field image can be digitally refocused to match a confocal microscopy image captured at a different sample plane.

Obesity and diabetes often lead to peripheral neuropathy. Damage and axonal die-back of the peripheral nervous system constitutes peripheral neuropathy. By 2030, half of the US adult population is projected to be obese, and type 2 diabetes mellitus is most commonly caused by obesity. As incidences of obesity and diabetes increase, the adverse effects of neuropathy will also increase. Neuropathy, previously thought to only affect skin layers of distal extremities, has recently been discovered in subcutaneous adipose tissue depots. Obese adipose tissue is fibrotic, resulting in excess collagen deposition. Collagen organizes the peripheral nervous system, but its interaction with adipose nerves has not been thoroughly investigated. Using 2-photon microscopy combined with second harmonic generation microscopy, we examined the spatial relationship between collagen and nerve in the adipose microenvironment to gain a better understanding of neuropathy pathways and mechanisms. Pearson’s Correlation Coefficient analysis suggests that an obese diet leads to greater colocalization between nerve and collagen in adipose tissue than a lean diet. These findings motivate further investigation as the Pearson Correlation Coefficient is restrictively optimized for structures that are overlapped, whereas nerves may simply be wrapped with or tightly associated with collagen. Here we present an adaptation of the multiscale 2D Wavelet Transform Modulus Maxima method to reveal different anisotropic signatures across adiposeresiding nerve and collagen fibers in tissues from mice fed obese and lean diets, respectively. Based on these promising preliminary results, additional development of multiscale wavelet-based techniques will offer insight into neuropathy through thorough investigation of nerve and collagen spatial relationships.

Conventional three-dimensional (3D) images of biological samples are typically assembled from a stack of twodimensional images acquired sequentially at different focal planes. This time-consuming manner hinders the application of 3D imaging techniques to the investigation of fast biochemical dynamics and light-sensitive biological events. The concept of multifocus imaging, which enables simultaneous acquisition of images from multiple focal planes, was introduced to achieve rapid 3D imaging. In the present study, we achieved multifocus imaging through polarization wavefront shaping via a micro-retarder array which splits the incident linearly polarized light into three beamlets that are focused to three axially-offset focal planes with ~100 μm separation. Append to an existing beam-scanning microscope, this multifocus system enables rapid 3D imaging compatible with a variety of optical microscopic approaches including laser transmittance, two-photon excited fluorescence, and second harmonic generation microscopy.

To capture an all-in-focus and 3D depth image, shape from focus method is widely used. The phase accuracy of the image candidates should be processed and adjusted beforehand. Phase only correction method was employed in this work and to accelerate the processing speed, the image processing by Fast Fourier Transform (FFT) was optimized. Meanwhile, the processing task was assigned in several parallel threads so that the performance would be improved. The method was used on a variable focus imaging system, and, as a result, the processing speed was improved to around 2.5-fps.

3D refractive index imaging methods usually rely on a weak-scattering approximation that does not allow for thick samples to be imaged accurately. Recent methods such as 3D Fourier ptychographic microscopy (FPM) instead use multiple-scattering models which allow for thicker objects to be imaged. In practice the illumination-side coding of 3D FPM requires redundant information and may produce inaccurate reconstructions for thick samples. Here, we propose augmenting 3D FPM with detection-side coding using a spatial light modulator (SLM) and optimize the SLM coding strategy with physics-based machine learned pupil coding designs that are optimized for 3D reconstructions. We compare our learned designs to random-, defocus-, Zernike aberrations-based pupil codes in simulated and experimental results.

Hyperspectral imaging (HSI) technology has been applied in a range of fields for target detection and mixture analysis. While its original applications were in remote sensing, modern uses include agriculture, historical document authentications and medicine. HSI has shown great utility in fluorescence microscopy; however, acquisition speeds have been slow due to light losses associated with spectral filtering. We are currently developing a rapid hyperspectral imaging platform for 5-dimensional imaging (RHIP-5D), a confocal imaging system that will allow users to obtain simultaneous measurements of many fluorescent labels. We have previously reported on optical modeling performance of the system. This previous model investigated geometrical capability of designing a multifaceted mirror imaging system as an initial approach to sample light at many wavelengths. The design utilized light-emitting diodes (LEDs) and a multifaceted mirror array to combine light sources into a liquid light guide (LLG). The computational model was constructed using Monte Carlo optical ray software (TracePro, Lambda Research Corp.). Recent results presented here show transmission has increased up to 9% through parametric optimization of each component. Future work will involve system validation using a prototype engineered based on our optimized model. System requirements will be evaluated to determine if potential design changes are needed to improve the system. We will report on spectral resolution to demonstrate feasibility of the RHIP-5D as a promising solution for overcoming current HSI acquisition speed and sensitivity limitations.

We have proposed virtual-phase-conjugation-based optical tomography (VPC-OT) using a virtual phase conjugation technology for single-shot and three-dimensional optical tomography. In VPC-OT, a random-spatial-phase-modulated probe beam is irradiated to the sample to be measured, and the complex amplitude of the signal composed of a superimposition of light reflected from each layer of the sample is measured. A three-dimensional tomogram of intensity and phase is obtained by reproducing the measured complex amplitude using a phase conjugate wave in a virtual optical system built in a computer. At this time, by changing the parameters of the virtual optical system, it becomes possible to obtain information of various tomographic planes from the data obtained with a single measurement. In the ideal virtual phase conjugate reproduction process, free space propagation can be assumed; however, in the actual measurement, due to the distortion of the waves and the surroundings of the sample to be measured, a mismatch will occur in modulation and demodulation, and the separation accuracy between different tomographic planes would be degraded. We perform an experiment to clarify the characteristics of VPC-OT in this situation. In this experiment, three-dimensional optical tomography is performed using an etching glass having a periodic structure of 30 μm as a sample, and the phase distribution is measured quantitatively. Furthermore, by placing a cover glass in front of the object and performing the same measurement, we discuss the characteristics and performance of VPC-OT when there is an optical distortion around the sample to be measured.

Combined with confocal imaging, Fluorescence lifetime imaging microscopy (FLIM) can achieve 3-dimensional optical sectional capability with sub-nanosecond lifetime information. As confocal FLIM acquires multi-dimensional data 4D (3D space + time), it is inherently slow. Recent developments in lock-in pixel imagers with time gated pixels show such detectors are capable of collecting as many as 8-time gates in a single pixel cycle. We present a multiplexed confocal FLIM microscope, equipped with a 4-taps time-gated lock-in pixel imager. The multiplexing setup allows the use of the sparse array with sub-nanosecond time-gating to achieve high throughput FLIM acquisition.

We propose the multiple focused inclined beams in conjunction a highly efficient detection system and electronically controllable confocal slit array in the fluorescence microscope system to obtain low photodamage, good optical sectioning, video-rate, and high-sensitivity imaging. To do so, we use a digital micro-mirror to generate the multiple beams and to form the detection slit synchronously. This provides the high resolution, the capability to image thick sample and the longer observation time while maintaining a similar imaging speed. By directly programming the digital micro-mirror we could correct the anisotropic beam shape and suppress the background.

In this work, several tests of the performance of 3,3’-thiodipropanol (TDP) as a mounting medium for fluorescence microscopy of biological samples were performed. Besides optical properties like the dispersion curve of TDP and the effect of the embedding medium on the fluorescence of commonly used dyes, the interaction with biological specimens, including the labeling of filamentous actin with fluorescent phalloidins, was tested: TDP showed to represent an interesting alternative to commercial mounting media. By mixing TDP with 2,2’-thiodiethanol (TDE), it was possible not only to fine tune the refractive index of the resulting solution, but to preserve the compatibility with fluorescent phalloidins.

We developed a nonlinear optical microscopy (NLOM) system with an extended field of view (FOV) of up-to 1.6 mm × 1.6 mm, employing a high numerical aperture (NA) and low magnification objective lens (20×, NA>0.9) maintaining submicron lateral resolution and <2 μm axial resolution, making the system a suitable candidate for high resolution imaging. An acquisition speed of 95 M samples per second was implemented with synchronized sampling of 1 voxel per optical pulse from a femtosecond laser source with a 95 MHz repetition rate, with the ability of scanning a 1.6 × 1.6 × 1.6 mm³ volume, with 12000 × 12000 × 2000 (× 4 channels), i.e., 1.152 Tera-voxels in total, capturing ~2.1 Terabyte of 16-bit raw data with 14-bit resolution in <53 minutes at 0.8 μm Z-steps, and maintaining a Nyquist-exceeded voxel-size, a Nyquist-exceeded volume-scanning speed and a Nyquist-exceeded line-scanning speed of <15 attoliter, >1288 μm³/ ms and >12 mm/ ms, respectively.

Gabor domain optical coherence microscopy (GD-OCM) is an emerging tomographic imaging technique that overcome the depth of focus (DOF) limit of high numerical objective lens by acquiring multiple cross-section images corresponded with different focal planes of the objective along an axial direction and combining them to form an invariant high resolution cross-section image. We previously reported an alternative processing method of GD-OCM called spectral fusing GD-OCM (SF-GD-OCM). SF-GD-OCM extracts in-focus information directly from acquired spectral interference signals. The advantage of SF-GD-OCM over the original GD-OCM is that the Gabor filtering and fusing can be implemented in an acquisition hardware instead of a host computer. Here, we have investigated the implementation of SF-GD-OCM by using FPGA frame grabber. The experimental setup of the acquisition system as well as the implementation of the spectral Gabor fusion in FPGA device have been developed and verified. Results show that the data processing of SF-GD-OCM in FPGA enabled acquisition device is possible and could lead to improvement in overall imaging speed of GD-OCM.

In this study, a combined two-color phase plate (CTPP) was designed for super-resolution microscopy based on upconversion fluorescence depletion (FD), fabricated, and evaluated. It is composed of two types of phase plates, a spiral phase plate and an annular phase plate. A two-color phase plate modulates the phase of the erase beam while maintaining the phase of the pump beam. SRM performed using the proposed CTPP is expected to enable super resolution in both the focal plane and in the optical axis direction. Despite its complex structure, a highly accurate CTPP was obtained by using the exposure and etching processes used in semiconductor manufacturing.