It is shown that nanosecond to picosecond fluorescence relaxation phenomena can be accessed for imaging after double pulse saturation excitation. This new technique has been introduced before as fluorescence lifetime imaging (DPFLIm) (Mueller et al, 1995). An OPA laser system generating ultra short, widely tunable, high power optical pulses provides the means for the selective excitation of specific fluorophores at sufficient excitation levels to obtain the necessary (partial) saturation of the optical transition. A key element in the developed method is that the correct determination of fluorescence relaxation times does allow for non-uniform saturation conditions over the observation area. This is true for the validation demonstration experiments reported here as well as for imaging applications at a later stage. Measurements on bulk solutions of Rhodamine B and Rhodamine 6G in different solvents confirm the experimental feasibility of accessing short fluorescence lifetimes with this technique. As only integrated signal detection is required no fast electronics are needed, making the technique suitable for fluorescence lifetime imaging in confocal microscopy, especially when used in combination with bilateral scanning and cooled CCD detection.

In PSAF (point spread autocorrelation function) imaging a fluorescence signal is generated from the interference response in the overlap region of two spatially shifted point spread functions (PSFs). It is experimentally demonstrated that a resolution improvement of approximately 30% can be realized in the case of axially shifted PSFs under high numerical-aperture (NA) conditions. A similar improvement in resolution is expected from numerical modeling for the lateral case. The presented technique can also be applied to the measurement of the effective point spread function itself in all three dimensions. It is found that the technique -- which in the latter case uses a bulk fluorescing solution -- is an excellent tool to access the apodization conditions of a practical optical system, such as a high-numerical aperture objective.

We describe a new form of light-efficient real time confocal microscope which requires neither a laser nor scanning in order to obtain a real time confocal image. The approach is based on aperture correlation techniques.

Near-field microscopy usually uses an evanescent wave as an illumination source on a sample. An enhanced evanescent wave can be obtained at the surface of a multilayer thin-film coating system due to the optical tunneling effect. It can improve the illumination power by several orders of magnitude in near-field microscopy. The enhancement of an evanescent wave is related to the thickness and refractive indices of the thin film layers as well as the incident angle and wavelength of the illumination source. In this paper, optimization of these parameters is studied in detail under resonance conditions, and some experimental results also are presented.

We report our work towards improving the resolution performance of a conventional Mirau correlation microscope (MCM) by operating in the UV. We have modified the Mirau interferometer with the introduction of new thin-film beamsplitters that are nonattenuative in the ultra-violet, and new interferometer structure that can be made to work with short working distance objectives. Along with special UV optics, the new MCM can be made to operate in the ultra- violet region. By operating at 248 nm mercury line, the expected resolution improvement in both transverse and depth directions should be nearly doubled. A fully working MCM UV system at 325 nm is described in detail, and demonstrates the feasibility for a 248 nm system once a high quality 248 nm objective becomes available.

Measurements of refractive index spatial distribution within optically transparent phase 3-D samples with computerized interferometric microscopes is proposed. For the reconstruction of the phase information the phase shifting interferometry is used. Interferometric computed-tomography microscopes are fully automated integrated systems that incorporate CCD camera, frame grabber, computed operated PZT mirror as well as intelligible software for interferogram processing, tomographic reconstruction and refractive index distribution are displayed in 2-D and 3-D, sections, isolines, etc. Experimental results for interferometric microscopy and tomography of blood cells are presented.

In this paper we investigate the use of a geometric phase- shifting (GPS) technique which allows us to convert conventional transmission or reflection differential interference contrast (DIC) microscopy into a quantitative mode. A phase-shifting algorithm is employed to extract the specimen phase gradient from the mixture of phase and amplitude information which is common in DIC. Fourier techniques are then used to recover the exact phase (i.e. optical path length variations) throughout the biological specimen viewed. In addition to this quantitative 'phase map,' we demonstrate that the GPS process simultaneously yields an 'amplitude-only' representation in which various absorption and transmission properties of the specimen are displayed as intensity variations in the image, similar to brightfield microscopy. These two resulting images can then be analyzed or further processed in a number of ways that are not possible with conventional DIC and which improve the microscopist's ability to correctly identify, interpret and measure features in the specimen.

We discuss a number for configurations for confocal microscopes and measure the extinction coefficient as a function of pinhole size. The introduction of a Babinet- Soleil compensator/polarizer combination is found to result in an extremely high extinction coefficient.

Confocal interference microscopy can be used to obtain phase information in a confocal microscope. Different methods of recovering phase are discussed. A fiber optic confocal interference microscope has been developed. Confocal interference can be used to investigate aberrations in a confocal microscope. Methods of profiling surfaces are reviewed. Depth profiling of stratified media is considered.

Experimental results are given for the optical sectioning characteristics of finite-sized, multiple-aperture arrays in brightfield direct-view microscopy. We also present a theoretical model for the brightfield direct-view microscope (DVM). This allows us to determine the optical sectioning strength for finite-sized, multiple-pinhole arrays with an arbitrary distribution of apertures. The theoretical model is modified, for experimental purposes, to take account of the presence of objective lens pupil shading. A comparison between experimental and theoretical results for the axial response of the DVM to a plane mirror specimen is presented. In particular, the effects of pinhole size, pinhole spacing and array geometry are investigated in detail with a veil to (1) achieving good optical sectioning characteristics and (2) maximizing the amount of light available for imaging. The implications of our results for practical systems as regards pinhole-array design and fabrication are also discussed.

We present an instrument that uses a novel technique to render topographical maps of microscopic reflecting surfaces in which depth resolution can be an order of magnitude or more better than the lateral resolution of the confocal microscope. In a confocal scanning reflected-light microscope, the illumination light coming out of the objective is a cone of light with its axis parallel to the optical axis and focused into a diffraction-limited spot. If the specimen is smooth within the size of the spot, then the laws of reflection apply. When the reflecting surface is perpendicular to the optical axis, the reflected light is a cone that coincides with the illumination light. If the reflecting surface is not perpendicular to the optical axis, the axis of the reflected cone of light is at an angle with respect to the optical axis equal to twice the tilt angle of the surface and oriented in the direction of the surface tilt. Because of this tilt, a portion of the back focal plane (BFP) of the objective does not receive any of the reflected light, while in the diametrically opposite side, the aperture stop blocks part of the reflected light. For a given illumination profile, the distribution of light on the BFP is determined by the tilt and orientation angles of the reflecting surface and thus can be used to recover these angles for each pixel in the specimen. These angles are then used to render a topographical map of the reflecting surface.

The analysis of the three-dimensional structure of tissue, cells and cellular constituents play a major role in biomedical research. Three-dimensional images, acquired by confocal fluorescence microscopes play a key role in this analysis. However, the imaging properties of these microscopes give rise to diffraction-induced blurring phenomena. These distortions hamper subsequent quantitative analysis. Therefore, restoration algorithms that invert these distortions will improve these analyses. We have tested the performances of the Richardson-Lucy, and the ICTM algorithm in a simulation experiment and found a strong dependency of their performances on the signal-to-noise ratio of the image. We propose a pre-filtering to reduce the noise in the image without hampering the object. We have applied a Gaussian filter and a median filter prior to the restoration, and compensate for this extra blurring of the Gaussian in the restoration procedure. We show how this pre- filtering improves the performance of the restoration algorithms. The experiments were performed on spheres convolved with a confocal point spread function and distorted with Poisson noise.

We have developed efficient image restoration algorithms for restoration of images that are acquired by conventional and confocal fluorescence microscopy. Assuming additive Gaussian noise or Poisson noise in the image and Gaussian or entropy prior distributions, functionals are formulated that must be minimized to obtain maximum a posteriori (MAP) and maximum likelihood (ML) estimations. We propose computationally efficient algorithms to find the solutions. The quality of the MAP restorations is determined largely by the choice of the regularization parameter, which determines the tradeoff between fitting and smoothing the solution. We propose a normalization method to ease the interactive choice of the regularization parameter if the variance of the noise is known. The performance of the algorithms was tested using simulated fluorescence conventional microscopy and fluorescence confocal laser scanning microscopy. Several error measures and quantitative measurements were used to evaluate the quality of the restoration result. We have tested the super-resolution capabilities and have found that the algorithms are capable of recovering partially the frequencies that were lost. The performance of the algorithms was compared to two existing algorithms that are commonly used for fluorescence imaging: the accelerated EM algorithm of Holmes and the regularized algorithm of Carrington.

All algorithms for three-dimensional deconvolution of fluorescence microscopical images have as a common goal the estimation of a specimen function (SF) that is consistent with the recorded image and the process for image formation and recording. To check for consistency, the image of the estimated SF predicted by the imaging operator is compared to the recorded image, and the similarity between them is used as a figure of merit (FOM) in the algorithm to improve the specimen function estimate. Commonly used FOMs include squared differences, maximum entropy, and maximum likelihood (ML). The imaging operator is usually characterized by the point-spread function (PSF), the image of a point source of light, or its Fourier transform, the optical transfer function (OTF). Because the OTF is non-zero only over a small region of the spatial-frequency domain, the inversion of the image formation operator is non-unique and the estimated SF is potentially artifactual. Adding a term to the FOM that penalizes some unwanted behavior of the estimated SF effectively ameliorates potential artifacts, but at the same time biases the estimation process. For example, an intensity penalty avoids overly large pixel values but biases the SF to small pixel values. A roughness penalty avoids rapid pixel to pixel variations but biases the SF to be smooth. In this article we assess the effects of the roughness and intensity penalties on maximum likelihood image estimation.

For the analysis of learning processes and the underlying changes of the shape of excitatory synapses (spines), 3-D volume samples of selected dendritic segments are scanned by a confocal laser scanning microscope. For a more detailed analysis, such as the classification of spine types, binary images of higher resolution are required. Simple threshold methods have disadvantages for small structures because the microscope point spread function (PSF) causes a darkening and a spread. The direction-dependent PSF leads to shape errors. To reconstruct structures and edge positions with a resolution smaller than one voxel a parametric model for the dendrite and the spines is created. In our application we use the known tree-like structure of the nerve cell as a- priori information. To create the model, simple geometrical elements (cylinders with hemispheres at the ends) are connected. The model can be adapted for size and position in sub-pixel domain. To estimate the quadratic error between the microscope image and the model, the model is sampled with the same resolution as the microscope image and convolved by the microscope PSF. During an iterative process the parameters of the model are optimized. In contrast to other pixel-based methods. the number of variable parameters is much slower. The influence of small deviations in the microscope image (caused by the inhomogeneous biological materials) is reduced.

A number of algorithms have been developed for three- dimensional (3D) deconvolution of fluorescence microscopical images. These algorithms use a mathematical-physics model for the process of image formation and try to estimate the specimen function, i.e. the distribution of fluorescent dye in the specimen. To keep the algorithms tractable and computational load practical, the algorithms rely on simplifying assumptions, and the extent to which these assumptions approximate the actual process of image formation and recording has a strong effect on the capabilities of the algorithms. The process of image formation is a continuous-space process, but the algorithms must be implemented using a discrete-space approximation to this process and render a sampled specimen function. A commonly-used assumption is that there is one pixel in the specimen for each pixel in the recorded image and that the pixel size in the recorded image is small compared to the size of the diffraction limited spot or Airy disk, a condition necessary to satisfy Nyquist sampling criterion. Modern CCD cameras, however, have large wells that integrate into a single pixel an area of the image that is significantly larger than the Airy disk. We derived a maximum-likelihood-based algorithm to accommodate for these large CCD pixel sizes. In this algorithm we assume that each pixel in the recorded image integrates several pixels that satisfy Nyquist criterion. The algorithm then attempts to estimate the specimen function at a resolution better than that allowed by the CCD camera. Preliminary results of this sub-pixel resolution algorithm are encouraging.

This work investigates a broad spectrum of conventional edge operators such as Sobel, Prewitt, Laplacian, Laplacian of Gaussian, and Roberts vis-a-vis upcoming multiresolution technique of wavelets. The technique of image decomposition with wavelets involves filtering of the image using a quadrature mirror filter pair (QMF) which brings out the details (using coefficients of highpass filter) and simultaneously smoothes the image (using coefficients of lowpass filter). Daubechies and Haar wavelet transforms have been used for the purpose of detecting edges. Haar wavelet has been found to be a most attractive choice for the purpose of edge detection with our methodology. A discussion and analysis of pros and cons of using wavelets against the conventional operators is presented in the paper.

A 780 nm stage-scanned confocal transmission microscope using compact disk (CD) player optics has been constructed and used to study test objects (including scattering effects) and breast tissue specimens (25-1500 micrometer thick). Confocal alignment in transmission is achieved using the CD optical mechanism moving coils to scan the objective lens until intensity peaks; the specimen is present but motionless. Intensity variation as a 3D function of mis- alignment of the two foci may allow estimation of optic deterioration caused by the specimen; appropriate confocal apertures and sample spacings can then be selected. Experimental measurements have been supplemented by 2D Monte-Carlo modeling of photon transport for a scattering- limited confocal transmission system. Modeling has characterized: transmission as a function of focal depth in a slab; contrast produced by a range of embedded spherical target radii; and optical mis-alignment. Poor dye perfusion within thick stained specimens results in little visible internal structure. This may be improved using partially phase dependent imaging (approximately equal to split- detector), sensitive to specimen refractive index variations.

The derivation of an iterative method for phase reconstruction from differential-interference-contrast (DIC) images is presented here. Because DIC imaging is direction sensitive, in our approach we estimate a specimen's phase function using multiple DIC images obtained by rotating the specimen. Results obtained from testing the method via two- dimensional simulations demonstrate that the use of multiple DIC images at different specimen rotations yield phase reconstructions that more closely resemble the phase function of phantoms than the unprocessed DIC images. Improvement in resolution was also achieved: two points separated by half the Rayleigh resolution limit for coherent illumination not resolved in the unprocessed DIC image, were successfully resolved in the reconstructed phase images. Our results show that phase reconstructions are quantitatively better and resolution is improved when two or more DIC images are used in the reconstruction.

We have combined Mie scattering theory and image theory to predict the forward scattering of light from spherical particles in a seeded fluid using high numerical aperture collection optics. Using this method, it is possible to determine all three components of a fluid's velocity by measuring the scattering from homogeneous spherical particles without moving the optics. The transverse velocity component is determined by following the centroid of the scattering pattern (with respect to time), while the component along the optical axis is determined by comparing the experimental data with numerical computations. We have verified our theoretical model and computer code by measuring the scattering from polystyrene particles illuminated with partially coherent, Koehler illumination in a transmitted light microscope. The three-dimensional scattering data is in quite good agreement with our model. To further verify our approach, we have measured the three- dimensional (parabolic) profile of a parallel flow of a low viscosity, seeded fluid in a straight channel (6 mm by 48 mm by 0.315 mm). The channel was placed on the stage of a conventional microscope equipped with a long working distance microscope objective, with the narrow dimension along the optical axis (OA). Instead of directly imaging the seed particles, the forward scattered light is recorded from the spherical, polystyrene seed particles (7 micrometer diameter).

This paper discusses some aspects of computer based modeling of biological microstructures. The workflow tom model and simulate a biological structure is described as a feedback- loop. Beside the system definition by structural and dynamical properties, the simulation is discussed as a mathematical representation coupled with a computer visualization. As an example, the investigation of the functional behavior of lung structures is described with special emphasis to the modeling of respiratory units.

A small depth of focus is very often a problem in the conventional high resolution light microscopy, because the objects are higher than the depth of focus. The connections of conventional light microscopes, video techniques and computers open new fields of applications for microscopy. It is possible to generate an image containing focused areas from a series of images of different focus- or z-positions. In this way it is possible to extend the depth of focus without a physical limitation of the numerical aperture (NA) of the objective lens. The connection of microscope, video and PC opens also the field of high resolution low light color applications, like fluorescence in situ hybridization (FISH). Image processing technology enables the enhancement of contrast for objects with a very low contrast. And such a system opens also the way to view objects with small motions, like processes of growing of biological objects. It is also the elimination of the limits in the X/Y-direction by a full color mosaics (patchwork mode) with a resolution up to 4500 by 3500 pixels. The last step in the development of the system was the generation of short movies of different 2D projections of the 3D data cube showing the spatial structure of the objects. It is possible to display up to 15 true color RGB-images per sec with a resolution of 664 by 512 pixels. Now the system is a powerful tool to generate image you never can see through the eyepieces.

Investigation of features and functions of some small biological objects (smaller than 500 nm), in particular, viruses, with conventional optical microscopy is practically impossible. Usually their images are obtained with methods of scanning electron microscopy (SEM), which precludes work with samples in a native state. We obtained images of different viruses including influenza A virus in native state with computer-aided phase microscope (CPM) Airyscan, in which an He-Ne laser is used as a light source. The main purpose of this work was to show the possibility to obtain adequate structure images of influenza viruses with diameter about 100 nm in conditions quite close to native and to investigate different stages of influenza virus budding. We suppose that these results may be considered as a basis for further studies of cell-virus interaction.

Solid state photon counters (SSPC), because of their small active area and high quantum efficiency, can offer a high quality, cost effective choice for scanned image acquisition. The speed and dynamic range of SSPCs show both thermal and electronic limitations. We present a series of experiments which separately examine the thermal and electronic saturation effects. We discuss several approaches to the thermal loading problem, demonstrate count rates of over 10⁷ counts/sec, and describe design considerations necessary for attaining rates of 10⁸ counts/sec.

A new algorithm is presented which uses maximum likelihood (ML) estimation and convex constraints to restore edge information in a robust and accurate way for microscope images. The convex constraints are spatially variant bounds on the image intensities and the directions of gradients, and they are extracted from an image restored with strong smoothing constraints. The high resolution estimate is obtained by maximizing the ML objective under the convex constraints and relaxed smoothing, using conjugate gradient and constrained optimization techniques. Relaxed smoothing allows edges in the final estimate, while the bound constraints preserve the robustness of the estimate. The convex constraints can be considered as a generalization of the positivity constraint which has been widely used in image restoration. The method does not impose any constraints on the profile of the edges, and it compares favorably with first order Bayesian methods which are frequently used for edge enhanced restoration in medical imaging. The performance of the method is demonstrated by restoring an image obtained from a confocal microscope.

Multi-photon (two or more photon) excitation imaging offers three significant advantages compared to laser-scanning confocal fluorescence microscopy for 3-D and 4-D fluorescence microscopy: considerable reduction in total sample excitation, increased depth penetration, and increased detection sensitivity. All-solid-state ultra-fast lasers offer tremendous potential for affordable, reliable, 'turn-key' multi-photon excitation sources. We have been developing a multi-photon system that utilizes an all-solid- state Nd:YLF excitation source. We have been evaluating the potential of this source for biological microscopy and have been optimizing system parameters for this application area. We have found that the 1047 nm radiation from these lasers can excite by two-photon fluorescence many commonly used fluorophores that are normally excited from blue to yellow light. In addition, we have found that this wavelength readily excites several normally UV excited fluorophores by the mechanism of three-photon excitation. The Nd:YLF laser has proven reliable in operation with nearly 6000 hours logged without significant loss of power. However, the original system produced rather long pulses for multi-photon excitation (300 fs) and a beam shape that was not ideal. We have recently commissioned the development of an improved pulse compressor from the manufacturers that gives narrower pulses (120 fs), improved beam shape, and a smaller insertion loss. This optimized excitation system has 6 times more potential two-photon excited fluorescence and 22 times more potential three-photon excited fluorescence than the prototype system. In addition, by optimizing coatings in the excitation and signal paths, we have improved the descanned detection sensitivity by 20% for two-photon excited fluorescence and 315% for three-photon excited fluorescence. The excitation optical transfer efficiency (1047 nm) of our imaging system is currently 60% to the back aperture of the objective. The emission optical transfer efficiency (670 nm) is currently 47% for descanned detection and 83% for non- descanned detection; both from the objective back aperture. Surprisingly, we find there is a signal-to-background increase of a factor of 17 between descanned and non- descanned modes of detection using a Nile Red solution sample.

Photobleaching causes progressive fading in successive slices of a through-focus series, so that the last image taken in the series can have a significantly lower signal- to-noise ratio than the first. Bleaching is often successfully minimized by including antifade additives in the specimen preparation and/or by reducing the optical dose to the specimen. However, these measures may not be sufficient in a through-focus series where many slices must be taken. This paper presents a simple approach to ameliorating the effects of bleaching, which is to progressively increase the integration times so as to maintain constant signal level from slice to slice. I refer to the sequence of integration times as an integration schedule. I develop the equations for integration scheduling from the physical assumptions and discuss how the method affects image quality.

We used a 10-micrometer-diameter fluorescent bead as a test object for 3D microscopy, and independently determined its structure by examining 1-micrometer-thick physical sections of the bead. Images of the full bead on different 3D microscopes revealed a number of aberrations and distortions. Images also showed evidence for absorption and/or scattering on all confocal and wide-field microscopes tested, but not on a two-photon microscope. The best 3D images of the bead came from deconvolution of either wide- field or certain confocal images. For deconvolution of partially confocal data, a region extending 7 micrometers beyond the top and bottom of the bead contained out-of- focus-light information essential for correct restoration. Fully confocal images required 10-fold less computer time for deconvolution, but 1000-fold more excitation light than a wide-field image, even though the resultant restorations were reasonably comparable. Laser imaging of the bead appeared to produce an artifactual image for which we currently have no explanation.