This PDF file contains the front matter associated with SPIE Proceedings Volume 7184, including the Title Page, Copyright information, Table of Contents, Introduction (if any), and the Conference Committee listing

One-dimensional linear detector arrays have been used in the development of microscopes. Our confocal line scanning microscope electronics incorporate two printed circuit boards: control board and detector board. This architecture separates control electronics from detection electronics allowing us to minimize the footprint at microscope detector head. The Field Programmable Gate array (FPGA) on the control board generates timing and synchronization signals to three systems: detector board, frame grabber and galvanometric mirror scanner. The detector is kept away from its control electronics, and the clock and control signals are sent over a differential twisted-pair cable. These differential signals are translated to single ended signals and forwarded to the detector at the microscope detector head. The synchronization signals for the frame grabber are sent over a shielded cable. The control board also generates a saw tooth analog ramp to drive the galvanometric mirror scanner. The analog video output of the detector is fed into an operational amplifier where the white and the black levels are adjusted. Finally the analog video is send to the frame grabber via a shielded cable. FPGA-based electronics offer an inexpensive convenient means to control and synchronize simple line-scanning confocal microscopes.

Both coherence and polarization play an important role in microscopy. It was long ago established that, in the context of scalar treatments, Köhler and critical illumination produce equivalent coherence functions in an object plane. This paper examines this equivalence in the context of polarization vortex illumination. Using a reversed-wavefront Young interferometer (RWYI), we show measurements of the coherence and correlation properties of the output illumination plane of various illumination systems. We compare the coherence properties of critical and Köhler vortex illumination schemes and look at the effects these properties have in an imaging system.

Fluorescence microscopy is an invaluable tool for studying biological processes in cells. In the recent past there has been significant interest in imaging cellular processes at the single molecule level. Single molecule experiments remove ensemble averaging effects and provide information that is typically not accessible through bulk experiments. One of the major requirements in single molecule imaging applications is that a sufficient number of photons be detected from the single molecule. This is not only important for the visual identification of single molecules, but also plays a crucial role in the quantitative analysis of the acquired data. Here, we demonstrate the use of a dual objective imaging configuration for single molecule studies. The configuration uses two opposing objective lenses, where one of the objectives is in an inverted position and the other objective is in an upright position. The use of opposing objective lenses has been previously demonstrated in 4pi confocal microscopy and I5M to achieve high axial resolution when compared to confocal/widefield microscopes. Here we demonstrate that the dual objective imaging configuration provides higher photon collection efficiency when compared to a regular microscope for a given illumination condition. As a result, single molecules can be localized with better accuracy when imaged through opposing objective lenses than when imaged through a regular optical microscope. Analytical tools are introduced to estimate the 2D location of single molecules and to characterize the accuracy with which they can be determined.

Phase microscopy has been widely used to acquire information about unstained transparent biological objects. These objects are well visualized by techniques such as Differential Interference Contrast (DIC) microscopy, but standard DIC systems don't provide quantitative phase information. Quantitative analysis is limited by the nonlinear relation between the intensity image and the object phase. In this work we combine nonlinearity concepts of the image formation process to model 3D DIC images, specifically the point spread function (PSF) formulation; capturing behavior of measured images along the optical axis that are not obtained when using other models. We verify our model by comparing simulated images to real measured images. This model is designed so that the reconstruction of the 3D properties of imaged specimens can be done in a simple form.

Several quantitative phase imaging techniques, such as digital holography, Hilbert-phase microscopy, and phase-shifting interferometry have applications in biological and medical imaging. Quantitative phase imaging measures the changes in the wavefront of the incident light due to refractive index variations throughout a 3-D specimen. We have developed a multimodal microscope which combines optical quadrature microscopy (OQM) and a Shack- Hartmann wavefront sensor for applications in biological imaging. OQM is an interferometric imaging modality that noninvasively measures the amplitude and phase of a signal beam that travels through a transparent specimen. The phase is obtained from interferograms with four different delayed reference wavefronts. The phase is then transformed into a quantitative image of optical path length difference. The Shack-Hartmann wavefront sensor measures the gradient of the wavefront at various points across a beam. A microlens array focuses the local wavefront onto a specific region of the CCD camera. The intensity is given by the maximum amplitude in the region and the phase is determined based on the exact pixel position within the region. We compare the amplitude and quantitative phase information of poly-methyl-meth-acrylate (PMMA) beads in oil and one-cell and two-cell mouse embryos with micrometer resolution using OQM and the Shack-Hartmann. Each pixel in OQM provides a phase measurement, whereas multiple pixels are used in Shack-Hartmann to determine the tilt. Therefore, the simple Shack-Hartmann system is limited by its resolution and field-of-view. Real-time imaging in Shack-Hartmann requires spatial averaging which smoothes the edges of the PMMA beads. The OQM has a greater field-of-view with good resolution; however, it is a complex system requiring multiple optical components and four cameras which may introduce additional artifacts in processing quantitative images. The OQM and Shack- Hartmann has certain advantages depending on the application. A combination of these two systems may provide improved quantitative phase information than either one alone.cHJl

In vitro fertilization (IVF) procedures have resulted in the birth of over three million babies since 1978. Yet the live birth rate in the United States was only 34% in 2005, with 32% of the successful pregnancies resulting in multiple births. These multiple pregnancies were directly attributed to the transfer of multiple embryos to increase the probability that a single, healthy embryo was included. Current viability markers used for IVF, such as the cell number, symmetry, size, and fragmentation, are analyzed qualitatively with differential interference contrast (DIC) microscopy. However, this method is not ideal for quantitative measures beyond the 8-cell stage of development because the cells overlap and obstruct the view within and below the cluster of cells. We have developed the phase-subtraction cell-counting method that uses the combination of DIC and optical quadrature microscopy (OQM) to count the number of cells accurately in live mouse embryos beyond the 8-cell stage. We have also created a preliminary analysis to measure the cell symmetry, size, and fragmentation quantitatively by analyzing the relative dry mass from the OQM image in conjunction with the phase-subtraction count. In this paper, we will discuss the characterization of OQM with respect to measuring the phase accurately for spherical samples that are much larger than the depth of field. Once fully characterized and verified with human embryos, this methodology could provide the means for a more accurate method to score embryo viability.

Double-helix point spread function (DH-PSF) is an engineered three-dimensional (3D) PSF specifically designed for 3D position estimation and imaging. It exhibits two lobes that rotate continuously around the optical axis with propagation. An information theoretical analysis shows that the DH-PSF carries higher and more uniform Fisher Information than the standard PSF through the 3D volume of interest. Experiments with DH-PSF demonstrate nanometer scale position localization accuracies in all three dimensions. Further, a variety of microscopy techniques such as bright-field, dark-field, and fluorescence can be directly transformed into their DH-PSF counterparts by placing an appropriate phase mask in the imaging path. The flexibility and performance make the DH-PSF attractive for 3D position localization in both photon-limited and photon-unlimited regimes.

Temporal image correlation provides a powerful fluorescence technique for measuring several biologically relevant parameters of molecules in living cells. These parameters include, but are not limited to local concentrations, diffusion dynamics, and aggregation states of biomolecules. However, the complex cellular environment presents several limitations, precluding high quantitative accuracy and constraining biological implementation. In order to address these issues, high speed spectral imaging was employed to compare the results of image correlation from spectrally unmixed and virtually implemented fluorescence emission filters. Of particular interest in this study is the impact of cellular autofluorescence, which is ubiquitous in fluorescence imaging of cells and tissues. Using traditional instrumentation, corrections for autofluorescence are commonly estimated as a static offset collected from a separate control specimen. While this may be sufficient in highly homogenous regions of interest, the low analyte concentrations requisite to fluctuation-based methods result in the potential for unbounded error resulting from spectral cross-talk between local autofluorescence inhomogeneities and the fluorescence signal of interest. Thus we demonstrate the importance of accurate autofluorescence characterization and discuss potential corrections using a case study focusing on fluorescence confocal spectral imaging of immune cells before and after stimulation with lipopolysaccheride (LPS). In these experiments, binding of LPS to the membrane receptor, YFP-TLR4, is observed to result in initiation of the immune response characterized by altered receptor diffusion dynamics and apparent heterogeneous aggregation states. In addition to characterizing errors resulting from autofluorescence spectral bleed-through, we present data leading to a deeper understanding of the molecular dynamics of the immune response and suggest hypotheses for future work utilizing hyperspectrally enabled multi-label fluorescence studies on this system of high biological import.

Confocal microscopes achieve high spatial resolution by focusing both a light source and a detector to a single point with an objective having a high numerical aperture. In order to produce an image, it is then necessary to scan this common focal spot through the specimen, and it is often important to image the full field of view in a short time. In order to avoid vignetting, the scanning must be done in or near the pupil of the optical system. For some fast scanners, this requires the use of multiple relay telescopes to create multiple pupils. Galvanometric scanners impose a practical limit on the scan speed because of the angular accelerations involved in reversing direction. Rotating polygons are often used to achieve greater speed. For a scanner consisting of a rotating polygon and a galvanometric scanner, two relay telescopes are normally used. We have developed a dual-wedge scanner which has the potential to perform the scan in a configuration which is very short in the axial direction, thereby eliminating much of the complexity of current high-speed scanners. We have demonstrated a prototype of the scanner in a reflectance confocal microscope. Transverse and axial resolution are comparable to those of other scanning systems. The selection of rotation speeds for the wedges is important to ensuring full coverage of the field of view in a reasonable time. Various tradeoffs on these parameters will be discussed. The beam behavior in the pupil will be discussed. Resolution limits and aberrations will be shown from ray-tracing analysis, and compared to experimental results.

We report a multi-wavelength digital holographic tomography system based on a fiber-based spectral interferometer. Instead of using tunable lasers, a relatively inexpensive broadband source was used as the light source. Hundreds of 2-D "synthesized holograms" (or object wavefields) were obtained by transversal scanning of a probe beam within a few seconds. Holographic images of an object volume were numerically reconstructed from each synthesized hologram and tomographic images were obtained by superposition of all the image volumes.

Digital Holographic Microscopy (DHM) allows quantitative multi-focus phase contrast imaging that has been found suitable for technical inspection and quantitative live cell imaging. The combination of DHM with fast and robust autofocus algorithms and a calibrated imaging system enables the determination of axial sample displacements. The evaluation of quantitative DHM phase contrast images permits also an effective detection of lateral object movements. Thus, data for 3D tracking is provided. Partially coherent light sources and multi-wavelength techniques open up prospects for an increased phase resolution in DHM by reduction of parasitic interference effects due to multiple reflections within the measurement setup. For this purpose, the utilization of light emitting diodes (LEDs) as well as the generation of short coherence properties by tunable laser light has been investigated for application in DHM. Results from investigations on sedimenting erythrocytes in suspension demonstrate that DHM enables (automated) quantitative dynamic 3D tracking of multiple cells without mechanical focus adjustment. Furthermore, it is shown that LEDs and multi-wavelength techniques enhance the axial resolution in inspection of reflective surfaces and quantitative digital holographic cell imaging.

Multiphoton excitation has recently found application in the fields of bioimaging, uncaging and lithography. In order to fully exploit the advantages of nonlinear excitation, in particular the axial resolution due to nonlinearity, most systems to date operate with point or multipoint excitation, while scanning either the laser beam or the sample to generate the illumination pattern. Here we combine the recently introduced technique of scanningless multiphoton excitation by temporal focusing with recent advances in digital holography to generate arbitrarily shaped, depth resolved, two-dimensional excitation patterns completely without scanning. This is of particular importance in applications requiring uniform excitation of large areas over short time scales, such as neuronal activation by multiphoton uncaging of neurotransmitters. We present an experimental and theoretical analysis of the effect of spatial patterning on the depth resolution achieved in temporal focusing microscopy. It is shown that the depth resolution for holographic excitation is somewhat worse than that achieved for uniform illumination. This is also accompanied by the appearance of a speckle pattern at the temporal focal plane. The origin of the two effects, as well as means to overcome them, are discussed.

High resolution wide-field imaging of the human retina calls for a 3D deconvolution. In this communication, we report on a regularized 3D deconvolution method, developed in a Bayesian framework in view of retinal imaging, which is fully unsupervised, i.e., in which all the usual tuning parameters, a.k.a. "hyper-parameters", are estimated from the data. The hyper-parameters are the noise level and all the parameters of a suitably chosen model for the object's power spectral density (PSD). They are estimated by a maximum likelihood (ML) method prior to the deconvolution itself. This 3D deconvolution method takes into account the 3D nature of the imaging process, can take into account the non-homogeneous noise variance due to the mixture of photon and detector noises, and can enforce a positivity constraint on the recovered object. The performance of the ML hyper-parameter estimation and of the deconvolution are illustrated both on simulated 3D retinal images and on non-biological 3D experimental data.

Several reports in the biological literature have indicated that when a living cell divides, the two daughter cells have a tendency to be mirror images of each other in terms of their overall cell shape. This phenomenon would be consistent with inheritance of spatial organization from mother cell to daughters. However the published data rely on a small number of examples that were visually chosen, raising potential concerns about inadvertent selection bias. We propose to revisit this issue using automated quantitative shape comparison methods which would have no contribution from the observer and which would allow statistical testing of similarity in large numbers of cells. In this report we describe a first order approach to the problem using rigid curve matching. Using test images, we compare a pointwise correspondence based distance metric with a chamfer matching strategy and find that the latter provides better correspondence and smaller distances between aligned curves, especially when we allow nonrigid deformation of the outlines in addition to rotation.

In this paper we use kymographs and computational image processing to convert 3-D video microscopy data of intracellular motion into 1-D time series data for further analysis. Because standard tools exist for time series analysis, this method allows us to produce robust quantitative results from otherwise visual data. The kymograph-based approach has an additional advantage over standard particle-tracking and flow-based image quantification algorithms in that we can average out camera noise over the spatial axis of the kymograph. The method has the disadvantage that it removes all spatial information. For this reason we see this method as a complement to rather than a replacement of standard tracking algorithms. The standard problem we are trying to address in our work is how fluorescent proteins in one cellular compartment are injected into another cellular compartment. The proteins travel at constant speed along a fixed spatial path, so a 2-D kymograph produced from a trace along this fixed path will tell us about the injection history into this second compartment. Our algorithm works by first taking a Radon transform of the input 2-D kymograph. We next make synthetic kymographs by backprojection. The angle with the best correlation between the original kymograph and the backprojection determines the dominant speed of the moving particles as well as the angle of the 1-D projected time series. Time series are then analyzed with standard tools to determine the peak size distribution, the peak interval distribution, the autocorrelation and the power spectrum.

This paper describes the application of a coded aperture snapshot spectral imager (CASSI) to fluorescence microscopy. CASSI records an interleaved spatially varying, spectrally filtered map of an object on a 2D focal plane. Convex optimization techniques combining least squares QR factorization with a total variance constraint are used to reconstruct a 3D data cube from a spectrally encoded 2D scene. CASSI records a 3D dataset at video rate - making it suitable for dynamic cellular imaging. We report on the reconstruction of fluorescent microspheres used in fluorescence microscopy applications and compare the results with images from a multi-spectral confocal system.

We present a versatile scheme for two-dimensional (2D) resolution enhancement in standing wave fluorescence microscopy (SWFM). This SWFM scheme consists of an interferometer, where both beams are focused at the back focal plane of the objective lens. Their position is controlled by a pair of acousto-optic deflectors (AODs). This results in two collimated beams that interfere in the focal plane, creating a lateral periodic excitation pattern with controlled fringe spacing and orientation. The phase of the standing wave (SW) pattern is controlled by the phase delay between two RF sinusoidal signals driving the AODs. An enlarged fluorescence image formed using the same objective lens is captured by a cooled CCD camera. Data collection involves acquiring images with excitation pattern of three equi-polar orientations (0°, 60° and 120°) and three different phases (0°, 120°, 240°) for each orientation. The SWFM image is algebraically reconstructed from these 9 acquired images. The SWFM image has enhanced 2D lateral resolution of about 100 nm with nearly isotropic effective point-spread function (PSF). As a result of the acousto-optic scanning, the total acquisition time can be as short as 100 μs and is only further limited by the fluorescence intensity, as well as sensitivity and speed of the CCD camera. Utilizing acousto-optic laser scanning for advanced SWFM provides the exceptional precision and speed necessary for real-time imaging of subresolution processes in living biological systems.

We present an alternative high-resolution fluorescence imaging technique, saturated excitation (SAX) microscopy, for observations of biological samples. In the technique, we saturate the population of fluorescence molecules at the excited state with high excitation intensity. Under this condition, the fluorescence intensity is no longer proportional to the excitation intensity and the relation of the fluorescence and excitation intensity shows strong nonlinearity. In the centre of laser focus, the nonlinear responses induced by the saturation appear notably because of higher excitation intensity. By detecting fluorescence signals from the saturated area, we can push the spatial resolution beyond the diffraction barrier in three dimensions. SAX microscopy can be realized with a simple optics, where a laser intensity modulation sisytem and a lock-in amplifier are simply added to a conventional confocal microscope system. Using the SAX microscope, we demonstrated high-resolution imaging of a biological sample by observing mitochondria in HeLa cells. We also examined the nonlinear response of commercially available dyes under saturated excitation conditions.

We used nonlinear fluorescence emission under the condition of saturated excitation (SAX) of fluorescent molecules for high-resolution laser scanning microscopy. In the technique, SAX microscopy, we modulate the excitation intensity at a single frequency and demodulate the fluorescence signal at a harmonic frequency to extract a nonlinear fluorescence response that contributes to improvement of the spatial resolution. This nonlinear fluorescent response on saturated condition was analyzed by rate equations formulated from a five-level system Jablonski diagram. By calculating relationship between excitation intensity and fluorescence signal demodulated at harmonic frequencies for rhodamine-6G molecules with 532 nm excitaion, we found that the fluorescent signal exhibits high-order nonlinear dependence on the excitation intensity under conditions of saturated excitation. We also calculated effective point spread functions (ePSFs) of SAX microscopy. The result of the calculation shows that ePSFs given with the harmonic demodulation provides the spatial resolution beyond the resolution limit of conventional confocal microscopy. The optical transfer functions have also been calculated from the ePSFs. The result of the calculation shows that a higher spatial-resolution can be obtained by demodulating fluorescnece signal at a higher harmonic frequency without theoretical limitation.

Linear Structured Illumination is a powerful technique for increasing the resolution of a fluorescence microscope by a factor of two beyond the diffraction limit. Previously this technique has only been used to image fixed samples because the implementation, using a mechanically rotated fused silica grating, was too slow. Here we describe a microscope design, using a ferroelectric spatial light modulator to structure the illumination light, capable of linear structured illumination at frame rates up to 11Hz. We show live imaging of GFP labeled Tubulin and Kinesin in Drosophila S2 cells.

Biological macromolecular interactions between proteins, transcription factors, DNA and other types of biomolecules, are fundamentally important to several cellular and biological processes. 3D Multi-channel confocal microscopy and colocalization analysis of fluorescent signals have proven to be invaluable tools for detecting such molecular interactions. The aim of this work was to quantify colocalization of the FOXP3 transcription factor in 3D cellular space generated from the confocal 3D image sets. 293T cells transfected with a conditionally active form of FOXP3 were stained for nuclei with Hoechst, for FOXP3 with anti-FOXP3 conjugated to PE, and 4-hydroxytamoxifen used as protein translocation and activation agent. Since the protein signal was weak and nonspecific intensity contributions were strong, it was difficult to perform colocalization analysis and estimate colocalization quantities. We performed 3D restoration by deconvolution method on the confocal images using experimentally measured point spread functions (PSFs) and subsequently a color shift correction. The deconvolution method eliminated nonspecific intensity contributions originating from PSF imposed by optical microscopy diffraction resolution limits and noise since these factors significantly affected colocalization analysis and quantification. Visual inspection of the deconvolved 3D image suggested that the FOXP3 molecules are predominantly colocalized within the nuclei although the fluorescent signals from FOXP3 molecules were also present in the cytoplasm. A close inspection of the scatter plot (colocalization map) and correlation quantities such as the Pearsons and colocalization coefficients showed that the fluorescent signals from the FOXP3 molecules and DNA are strongly correlated. In conclusion, our colocalization quantification approach confirms the preferential association of the FOXP3 molecules with the DNA despite the presence of fluorescent signals from the former one both in the nuclei and cytoplasm.

We present the method measuring the thickness and the refractive index of a transparent specimen at a same time based on full-field optical coherence tomography. As a sample a small drop of epoxy was placed on a flat plate and the high-resolution depth resolved en-face images of the epoxy drop were taken. With adopting the plate surface as a reference plane, the physical thickness and the refractive index distribution could be obtained. Owing to the full-field imaging capability, we could obtain the transverse distributions of the thickness and the refractive index without any transverse scanning. The measured thickness at the center of the sample was 24 μm and the average index was 1.4055 with the standard deviation of 0.0002.

Fluorescence lifetime imaging microscopy is a technique in which the fluorescence lifetime(s) of a fluorophore is measured at each spatially resolvable element of a microscope image. Imaging of fluorescence lifetimes enables biochemical reactions to be followed at each microscopically resolvable location within the cell. FLIM has thus become very useful for biomedical tissue imaging. Global analysis [1] is a method of recovering fluorescence decay parameters from either time-resolved emission spectra to yield Decay-Associated Spectra [2], or equivalently, from FLIM datasets to yield Decay-Associated Images. Global analysis offers a sensitive and non-invasive probe of metabolic state of intracellular molecules such as NADH. Using prior information, such as the spatial invariance of the lifetime of each fluorescent species in the image, to better refine the relevant parameters, global analysis can recover lifetimes and amplitudes more accurately than traditional pixel-by-pixel analysis. Here, we explain a method to analyze FLIM data so that more accurate lifetimes and DAIs can be computed in a reasonable time. This approach involves coupling an iterative global analysis with linear algebraic operations. It can be successfully applied to image, e.g. metabolic states of live cardiac myocytes, etc.