A system for quantification of neurite outgrowth in in-vitro experiments is described. The system is developed for routine use in a high-throughput setting and is therefore needs fast, cheap, and robust. It relies on automated digital microscopical imaging of microtiter plates. Image analysis is applied to extract features for characterisation of neurite outgrowth. The system is tested in a dose-response experiment on PC12 cells + Taxol. The performance of the system and its ability to measure changes on neuronal morphology is studied.

A complete system for object segmentation, counting, quantification, and tracking from microscopic images was implemented. We found that image deconvolution and reconstruction operations are essential to the success of any general-purpose segmentation algorithm and hence are of paramount importance for a counting and tracking software system. Wavelet-based image enhancement, background equalization, and noise suppression routines are the components in our novel general-purpose segmentation algorithm. Simple object recognition based on averages and preset tolerances suffices for most applications. As expected, boundary smoothing is important if watershed-based blob separation is to be used. One of the challenges of a general-purpose counting and tracking system is the need for a large number of object quantification components (features). In tracking we found that incorporating weighted features into an error function improves the accuracy over just the path coherence criterion and that evaluating correspondences over multiple time frames improves the accuracy over using only two consecutive time frames.

3-D images obtained from optical sectioning microscopy are usually degraded by a point-spread function that is known to be an even function but is otherwise only approximately known, or even entirely unknown. We present a new algorithm for 3-D blind deconvolution of even point-spread functions that is both fast and (in the absence of noise) exact. Fourier transforms decouple the problem into 2-D, then 1-D blind deconvolution problems, greatly increasing computational speed. Numerical simulations demonstrate that the blind algorithm seems to perform both faster and more accurately than the non-blind iterative Lucy-Richardson algorithm.

A reconstruction algorithm is developed that uses specific a-priori knowledge to produce higher resolution images than standard approaches. Deconvolution is an important image reconstruction tool in fluorescence microscopy. This is especially true for modern interferometric instruments (such as I5M and 4Pi systems), as they may have complicated oscillatory point spread functions. Current methods are designed to work on an arbitrary object - i.e. it is assumed that there is no available a-priori knowledge of the object (with the possible exception of a non- negative condition on the fluorophore-emission intensities). In situations where there is a-priori knowledge of the object, it may be possible to use this information to produce a higher quality reconstruction of the object. A useful a-priori condition is investigated here. It is assumed that the object can be represented by the sum of not more than L basis functions. The simplest example of this is when the basis functions are impulses - this leads to an object of L or less non-zero points on a background of zeros. This a-priori condition can be applied directly; applied to a limited region of the object; applied in one dimension (for an object with a layered structure such as lipid bilayers); or applied in two dimensions (for an object with a filamentary structure such as actin fibers.) A reconstruction algorithm is described and applied to some illustrative simulated examples. The results are found for several fluorescence microscopy methodologies and compared to the results produced by standard deconvolution methods.

The application of conventional confocal microscopes with high numerical aperture (NA) to in vivo imaging is limited by the objectiveís large physical dimensions and short working distance. We are developing a confocal microscope that uses simple low NA lenses oriented in a dual axes configuration for miniaturization and in vivo imaging. This architecture achieves a long working distance, micron level axial resolution, and reduced noise from scattered light outside the focal volume. Combined with the novel method of post-objective scanning, this design can be scaled down to millimeter dimensions. We derive the dual axes response from diffraction theory, and construct two tabletop prototypes to demonstrate the performance of this approach. We collect images from freshly excised biopsy specimens of human esophagus and transgenic mouse cerebellum expressing GFP. With horizontal cross-sectional images, we achieve 1 to 2 μm resolution and collect reflectance and fluorescence images. With vertical cross-sectional images, we achieve 4 to 5 μm resolution, dynamic range of 70 dB, and tissue penetration over 1 mm. An instrument miniaturized with this configuration could be used for in vivo cellular and molecular imaging.

Neurons are known to possess active computational properties. To investigate these properties, it is desirable to study the electrical and chemical properties not only at a living neuron's cell body, but also at many sites within its dendritic arborization. However, currently available recording techniques force a tradeoff between spatial and temporal resolution. To overcome these limitations, we have developed a confocal microscope that can make multisite optical recordings at an effective frame rate that is sufficient to measure fast neuronal events, such as action potentials, that occur on a timescale of milliseconds. We accomplished this by combining acousto-optic deflectors for addressable point illumination with a digital micromirror device for addressable point detection. After developing a registration algorithm to ensure synchronicity between point illumination and point detection, we used light-scattering test preparations to demonstrate that our system is capable of optical sectioning and therefore capable of imaging in living brain tissue. Furthermore, we have shown that fluorescence changes can be monitored at an effective frame rate of 25 kHz.

Confocal microscopy has a unique optical sectioning property which allows three-dimensional images at different depths. Use of a microlens array is a potential alternative to the Nipkow disk for parallel imaging with high throughput in real-time confocal microscopy. The use of variable-focal-length microlenses can provide a way to axially scan the foci electronically avoiding the inflexible mechanical movement of the lens or the sample. Here we demonstrate a combination of a variable-focal-length microlens array and a fiber optic bundle as a way to create a high throughput aperture array that would be potentially applied as confocal imaging in vivo biological specimens. Variable focal length microlenses that we use consist of a liquid crystal film sandwiched between a pair of conductive substrates with patterned electrodes. The incident side of the microlens array was determined by examining the focus distribution in the axial direction. The variation of the focal length obtained by changing the voltage and corresponding focus intensity were measured through a conventional microscope. Meanwhile, the fiber bundle was characterized by coupling with either coherent or incoherent light source. We use the fiber bundle as both a multiple aperture and an image-carrying element and combine it with a microlens array to built up a confocal system. Axial responses are measured in two optical arrangements as a route to investigate endoscope potential.

The advent of Electron Multiplying Charge Coupled Device (EMCCD) technology and it's ability to overcome previous hurdles in low-light fluorescence microscopy, such as phototoxicity to live cells, photobleaching of fluorophores and exposure time restrictions, has resulted in a significant resurgence of interest in use of confocal spinning disk techniques for live cell microscopy. Here provide an understanding of, and technical solutions to, the issues of synchronization that have previously marred the coupling of fast CCD camera technology to confocal spinning disk arrangements. We examine the challenges arising from both old and new models of the Nipkow spinning disk confocal unit and suggest solutions throughout based on a sound comprehension of both (a) relative scan/exposure times; (b) relative orientation of the coupled devices; (c) optimisation of EMCCD clocking parameters.

Ideally, a no-moving parts fast and agile scanning confocal microscope system is required that can produce true real-time 3-D scans with precision and repeatability. In this paper, such agile optical confocal microsopy designs are proposed that enable high speed precise non-invasive 3-D imaging. These compact confocal microscopes can provide real-time pin-point focussed imaging to enable confocal slices in-vivo, thus greatly reducing motion artifacts. These microscopes can be modified into interferometric microscopes for phase contrast imaging. The proposed microscopes can also greatly improve confocal fluorescence imaging as needed for cancer detection. An ultracompact confocal probe tip connected to a single ultra-thin fiber is another design option allowing flexibility for usage in tight cavities.

We describe a simple method to produce arbitrary complex optical fields using a computer generated binary phase hologram in a 4-f optical.

The digital holographic microscopy is being developed, to offer a number of significant advantages and capabilities in phase imaging. The direct availability of both the amplitude and phase information offers a range of versatile processing techniques that can be applied to the complex field data, including the phase imaging which is particularly straightforward in digital holography. One of the techniques we are developing addresses the problem of 2-pi phase discontinuity in the phase image, where most of the conventional phase unwrapping algorithms require subjective intervention for multi-lambda discontinuity. In the present system, we generate two phase maps by digital holography using two different wavelengths. Numerical superposition of the two phase maps results in a new phase map whose effective wavelength is inversely proportional to the difference of the two wavelengths. The axial range can in principle be arbitrarily large compared to the wavelength, while maintaining axial resolution to a fraction of wavelength. We apply the phase imaging digital holography to a number of systems, including the imaging of thin films and biological cells. Preliminary results from these experiments are presented and future development and applications are discussed.

We are developing the digital interference holography (DIH) as a novel method of microscopic tomographic imaging by numerical superposition of a number of holographic fields taken with varying wavelengths. The digital interference holography does not involve pixel-by-pixel mechanical scanning of three-dimensional volume and yet achieves high lateral and longitudinal resolutions. The holographic interference pattern of an object is generated optically and recorded digitally using a CCD camera. The hologrpahic image field is numerically calculated using basic diffraction formulas and the process is repeated for a range of varying wavelengths at regular intervals. Numerical superposition, or digital interference, of the holographic image fields yields the desired three-dimensional representation of the object. Experiments have demonstrated a few-micron lateral and axial resolutions. Furthermore, since the DIH is a coherent imaging system, one can form true tomographic images of sub-surface structures, in the presence of diffuse scattering from overlying layers. By being able to generate true tomographic images of subsurface structures, without the need for three-dimensional mechanical scannig, the DIH method can provide a very efficient and versatile imaging modality for a wide range of applications in materials science and biomedical imaging.

We present a multispectral polarimetric imaging system well suited for complete Mueller matrix microscopy. The source is a spectrally filtered halogen light bulb, and the image is formed on a fast CCD camera The light polarization is modulated before the sample and analyzed after the sample by using nematic liquid crystal modulators.. The whole Mueller matrix image of the sample is typically measured over 5 seconds for a good signal-to-noise ratio. The instrument design, together with an original and easy-to-operate calibration procedure provides a high polarimetric accuracy over wide ranges of wavelengths and magnifications. Mueller polarimetry provides separate images of scalar and vector retardation and dichroism of the sample, together with its depolarizing power, while all these effects do contribute simultaneously to the contrasts observed in standard polarized microsopy. Polarimetric images of several samples, namely an unstained rabbit cornea, a picrosirius red stained hepatic biopsy, and a rat artery specifically stained for collagen III are shown and discussed

Recently surgery requires extensive support from imaging technologies in order to increase effectiveness and safety of operations. One of important tasks is to enhance visualisation of quasi-phase (transparent) 3d structures. Those structures are characterized by very low contrast. It makes differentiation of tissues in field of view very difficult. For that reason the surgeon may be extremly uncertain during operation. This problem is connected with supporting operations of inner ear during which physician has to perform cuts at specific places of quasi-transparent velums. Conventionally during such operations medical doctor views the operating field through stereoscopic microscope. In the paper we propose a 3D visualisation system based on Helmet Mounted Display. Two CCD cameras placed at the output of microscope perform acquisition of stereo pairs of images. The images are processed in real-time with the goal of enhancement of quasi-phased structures. The main task is to create algorithm that is not sensitive to changes in intensity distribution. The disadvantages of existing algorithms is their lack of adaptation to occuring reflexes and shadows in field of view. The processed images from both left and right channels are overlaid on the actual images exported and displayed at LCD's of Helmet Mounted Display. A physician observes by HMD (Helmet Mounted Display) a stereoscopic operating scene with indication of the places of special interest. The authors present the hardware ,procedures applied and initial results of inner ear structure visualisation. Several problems connected with processing of stereo-pair images are discussed.

Few models, based on the diffraction theory, are proposed in order to evaluate the point spread function of different microscope objectives used in a digital holographic microscope. Because in holography the phase information is essential, a 3D amplitude point spread function (APSF), modulus and phase, is necessary, in order to properly deconvolute the 3D images obtained. Scalar Debye theory, paraxial approximation and vectorial Debye theory are used to solve the diffraction problem and the theoretical predicted 3D APSFs obtained with these models are compared.

Chondrocytes, obtained from testosterone treated human articular cartilage, were examined by a recently developed Multiple Beam Interference Microscopy (MBIM) attached to a confocal set up, Video-enhanced differential interference microphotography and also by cinematography. In the MBIM, the intensity of the transmitted pattern is given by the Airy function which increases the contrast dramatically as the coefficient of the reflectance of the parallel plates increases. Moreover, in this configuration, the beam passes several times through a specific organelle and increases its optical path difference both because of the increase in the trajectory and refractive index (high density) of the organelle. The improved contrast enhances the resolving power of the system and makes visible several structural details of sub micron dimensions like nucleolus, retraction fibers, podia, etc. which are not possible to reveal with such a clarity by conventional techniques such as bright field, phase contrast or DIC. This technique permits to detect the oscillatory and rotational motions of unstained cilia for the first time. The frequency of oscillations was found to be 0.8 Hz.

We present a high-resolution imaging technique using nanometric beads as multiple scattering local probes. The positions of the beads are determined in three dimensions by white-light interference microscopy. The technique has been applied to study the deformation of gels under mechanical constraint. The location of Brownian moving beads has also been demonstrated with nanometer spatial precision and 10 μs acquisition time. High-resolution 3-D imaging of hollow structures explored by the beads in relatively transparent materials should be possible.

We present a real-time, multi-dimensional, digital, optical coherence tomography (OCT) acquisition and imaging system. The system consists of conventional OCT optics, a rapid scanning optical delay (RSOD) line to support fast data acquisition rates, and a high-speed A/D converter for sampling the interference waveforms. A 1M-gate Virtex-II field programmable gate array (FPGA) is designed to perform digital down conversion. This is analogous to demodulating and low-pass filtering the continuous time signal. The system creates in-phase and quadrature-phase components using a tunable quadrature mixer. Multistage polyphase finite impulse response (FIR) filtering and down sampling is used to remove unneeded high frequencies. A floating-point digital signal processor (DSP) computes the magnitude and phase shifts. The data is read by a host machine and displayed on screen at real-time rates commensurate with the data acquisition rate. This system offers flexible acquisition and processing parameters for a wide range of multi-dimensional optical microscopy techniques.

The Micro-Optical Projection Tomographic Microscope (μOPTM) is an instrument that is being developed for three-dimensional (3D) imaging of cells and subcellular components. The target application for the μOPTM is the early detection of lung cancer by revealing the complex 3D information about chromatin redistribution in the nucleus. The µOPTM employs a scanning objective lens (100x, N.A.=1.25) to create an extended depth-of-field image, similar to a shadowgram or projection, that we call a pseudo-projection. A large number of pseudo-projections (90+) are acquired, from which a tomographically reconstructed 3D image is computed using a filtered backprojection algorithm. The prototype μOPTM uses a single objective lens, so the object (cell) must be rotated for each new pseudo-projection. A custom microtube stage minimizes the lateral and axial motion of the sample tube during scanning and rotation so that registration between successive pseudo-projections is maintained. Image processing methods are used to correct any remaining registration errors. The media inside and outside the tube are refractive index-matched to each other and to the tube (Δnavg < 0.02). The index-matched materials are pressed between two flat parallel windows, providing a nearly distortion-free image. Custom phantoms using microspheres have been constructed and images of these 3D test targets acquired. The minimum resolvable feature size is better than 3 microns. The first 3D image of a cell using μOPTM is also shown.

A computed-tomography microscope using a digital spatial light modulator has been developed and has demonstrated cytological and histological three-dimensional (3-D) image acquisition. The system consists of two high NA objective lenses, a Digital Micromirror Device (DMD) placed conjugate to the back pupil plane of the illumination objective, a sample stage, a light source, and a CCD detector. Each DMD micromirror can control the illumination of a specific angle. A single 3-D reconstruction is obtained from parallel ray projections acquired by changing the polar angle of illumination -phi_max < phi < phi_max while holding the azimuthal angle theta constant. The polar angle is limited by the NA of the objective. To compensate for an incompleteness of information due to the limited polar angles, several reconstructions acquired at multiple azimuthal angles are combined to create a final reconstruction. A reconstruction algorithm was developed using simulation software based on the 3-D Radon transformation and 3-D synthetic objects. Microscopic 3-D volume reconstructions of quantitatively absorption-stained cells have been demonstrated. 3-D reconstructed images enables the analysis of cell morphology and tissue architecture, as well as virtual two-dimensional slices with the distance between slices of 0.3 μm.

Experimental verification of our previously proposed linear phase imaging technique for differential inference contrast microscopy (DIC) microscopy is presented. This technique first applies phase-shifting methods to DIC to acquire linear phase gradient images in two orthogonal directions. A special Fourier integration algorithm is then applied to the combined phase gradient images to create a single linear phase image in which intensity is proportional to phase. This overcomes the limitations of traditional DIC, which cannot accurately measure the phase (i.e. refractive index or thickness) of embedded 3D phase objects. The linear phase imaging technique is implemented using a standard DIC microscope altered to allow controlled phase shifting, a low noise CCD camera, and post-processing in Matlab. The results presented confirm the linear proportionality of intensity to phase in these images.

Traditional light sources for fluorescence microscopy have been mercury lamps, xenon lamps, and lasers. These sources have been essential in the development of fluorescence microscopy but each can have serious disadvantages: lack of near monochromaticity, heat generation, cost, lifetime of the light source, and possible distortions due to coherence effects. We are examining the possibility of using the new high-power LED light sources as alternatives to the above mentioned sources. LED sources are near monochromatic, are inexpensive, produce little heat, have no coherence problems, have extended lifetimes, are small, and can easily be modulated. In this presentation we will describe experiments comparing various LEDs to other light sources. We will compare, for example, a 530 nm LED to the 546 nm line from a mercury lamp on a fluorophore whose absorption maximum is broad and in the middle between these two wavelengths.

Stereomicroscopy is an important method for use in image acquisition because it provides a 3D image of an object when other microscopic techniques can only provide the image in 2D. One challenge that is being faced with this type of imaging is determining the top surface of a sample that has otherwise indistinguishable surface and planar characteristics. We have developed a system that creates oblique illumination and in conjunction with image processing, the top surface can be viewed. The BFST consists of the Leica MZ12 stereomicroscope with a unique attached lighting source. The lighting source consists of eight light emitting diodes (LED's) that are separated by 45-degree angles. Each LED in this system illuminates with a 20-degree viewing angle once per cycle with a shadow over the rest of the sample. Subsequently, eight segmented images are taken per cycle. After the images are captured they are stacked through image addition to achieve the full field of view, and the surface is then easily identified. Image processing techniques, such as skeletonization can be used for further enhancement and measurement. With the use of BFST, advances can be made in detecting surface features from metals to tissue samples, such as in the analytical assessment of pulmonary emphysema using the technique of mean linear intercept.

Randomly selected pathology sections of lung tissue are used to correlate lung pathology with Computer Tomography (CT) images. The randomly selected pathology sections provide physicians with little freedom to thoroughly investigate specific areas of interest as identified via CT images. A Large Image Microscope Array (LIMA) was designed to serially section and image entire organs for direct correlation between lung pathology and CT. The LIMA consists of a novel vibratome, capable of sectioning tissue down to a thickness of 40mm at specimen dimensions of 20cm by 30cm to a total depth of 30cm. A camera and a stereomicroscope, mounted on a XYZ gantry above the vibratome is moved through an automated raster scan to capture the entire surface area of the tissue via many high magnification images. A custom software program was developed to automate all hardware components. The alignment and stitching of the images is achieved though custom C++ code in conjunction with the Insight Segmentation and Registration Toolkit (ITK). The resulting high magnification, high-resolution pathology images are registered with corresponding CT images. Through point-to-point correlation between the two imaging techniques a pathological and CT ground truth may be established.

We describe the design and the implementation of reflection confocal scanning microscopy (CSM) using an acousto-optical deflector (AOD) for the fast horizontal scan and a galvanometer mirror (GM) for the slow vertical scan. In the beam scanning system it is important to maintain the lateral and the axial performance during scanning operation. We propose a simple method to design a scanning system using the finite ray tracing and the diffraction theory. We define a cost function which contains the effect of aberrations on the performance of microscopy. We construct the designed system and evaluate its performance. The OSLO simulation shows that the performances of CSM are not changed with deflection angle. So we conclude that the beam scanning system is properly designed. In addition, we propose an image formation method and show images obtained with the system.

The errors can cause the serious loss of the performance of a precision machine system. In this paper, we propose the method of allocating the alignment tolerances of the components and apply this method to Confocal Scanning Microscopy (CSM) to get the optimal tolerances. CSM uses confocal aperture, which blocks the out-of-focus information. Thus, it provides images with superior resolution and has unique property of optical sectioning. Recently, due to these properties, it has been widely used for measurement in biological field, medical science, material science and semiconductor industry. In general, tight tolerances are required to maintain the performance of a system, but a high cost of manufacturing and assembling is required to preserve the tight tolerances. The purpose of allocating the optimal tolerances is minimizing the cost while keeping the performance of the system. In the optimal problem, we set the performance requirements as constraints and maximized the tolerances. The Monte Carlo Method, a statistical simulation method, is used in tolerance analysis. Alignment tolerances of optical components of the confocal scanning microscopy are optimized, to minimize the cost and to maintain the observation performance of the microscopy. We can also apply this method to the other precision machine system.

A new time-domain two-dimensional fluorescence lifetime detection method and a fluorescence lifetime imaging microscope that is based on a synchroscan streak camera are presented. This system can generate in parallel a complete two-dimensional fluorescence lifetime image without scanning the light. We demonstrate the acquisition of a fluorescence lifetime image of a semiconductor ET material sample at a temporal resolution of 2ps and spatial resolution of 4μm within 40ms.

The optical resolution limits the observation of fine features with size below diffraction limit, e.g. 500nm. We propose and demonstrate a technique that uses collimated laser scattering to characterize such features. We use a new illumination technique that let us image bio-molecules down to 50nm with optical microscopy. We observed vesicles with size down to 50nm and visualize the flow of such tiny molecules in the live cells. We could also characterize the detailed size and molecular weight of such particles.

Different sub-cellular compartments and organelles, such as cytosol, endoplasmic reticulum and mitochondria, are known to be differentially involved in Ca2+ homeostasis. It is thus of primary concern to develop imaging paradigms that permit to make out these diverse components. To this end, we have constructed a complete system that performs multi-functional imaging under software control. The main hardware components of this system are a piezoelectric actuator, used to set objective lens position, a fast-switching monochromator, used to select excitation wavelength, a beam splitter, used to separate emission wavelengths, and a I/O interface to control the hardware. For these demonstrative experiments, cultured HeLa cells were transfected with a Ca2+ sensitive fluorescent biosensor (cameleon) targeted to the mitochondria (mtCam), and also loaded with cytosolic Fura2. The main system clock was provided by the frame-valid signal (FVAL) of a cooled CCD camera that captured wide-field fluorescence images of the two probes. Excitation wavelength and objective lens position were rapidly set during silent periods between successive exposures, with a minimum inter-frame interval of 2 ms. Triplets of images were acquired at 340, 380 and 430 nm excitation wavelengths at each one of three adjacent focal planes, separated by 250 nm. Optical sectioning was enhanced off-line by applying a nearest-neighbor deconvolution algorithm based on a directly estimated point-spread function (PSF). To measure the PSF, image stacks of sub-resolution fluorescent beads, incorporated in the cell cytoplasm by electroporation, were acquired under identical imaging conditions. The different dynamics of cytosolic and mitochondrial Ca2+ signals evoked by histamine could be distinguished clearly, with sub-micron resolution. Other FRET-based probes capable of sensing different chemical modifications of the cellular environment can be integrated in this approach, which is intrinsically suitable for the analysis of the interactions and cross-talks between different signaling pathways (e.g. Ca2+ and cAMP).

A preliminary result supports the feasibility of using visible light to modulate the membrane potential of a cell. Human embryonic kidney cells (HEK293) were transfected with vertebrate rhodopsin and a gradient inward rectifying potassium (GIRK) channel. Whole cell patch clamp recordings of HEK293 cells exposed to 9-cis retinal showed that illumination increases the potassium current compared with recordings obtained in the dark. When combined with a rapid scanning device, this approach has the potential to control the activity of many neurons.