By utilizing lifetime information, in addition to spectral properties, new and interesting possibilities can be realized in confocal fluorescence microscopy. In combination with the previously described Intensity-modulated Multiple-beam Scanning (IMS) technique it is possible to strongly increase the number of fluorophores that can be simultaneously and independently recorded. Another possibility is to detect, simultaneously, changes in the lifetimes of multiple fluorophores caused by changes in the chemical environment such as pH or ion concentration. A factor that critically determines the usefulness of lifetime information is the signal-to-noise ratio (SNR) that can be attained. This paper presents a theoretical investigation of the SNR when using the IMS in combination lifetime measurements. It is found that when increasing the number of simultaneously detected fluorophores, it is important to select fluorophores with widely different lifetimes. For lifetime measurements the precision can be expressed in terms of the number of detected photons.

In order to separate different proteins, liquid chromatography is often used. The sample is pumped through a column filled with microspheres. The velocity of the proteins are depending on their interaction with the microspheres. The proteins could be labelled with a fluorescent marker and the distribution of the protein within the sphere can be recorded using a CSLM. When collecting optical sections using a CSLM the detected intensity decreases the deeper in the specimen the section is collected. This is due to absorption, scattering and bleaching. For the special case of a single microsphere it is of interest to find out how this combined effect is distributed within the sphere for a certain distribution of the fluorescent stain. When this distribution is known the attenuation can be compensated for. In the simulation the distribution of the stain is supposed to be the result of a diffusion process and all attenuation is supposed to arise from absorption only. The attenuation for a certain volume element (voxel) is supposed to occur from absorption in the voxels above, within the cone formed by the focused excitation light beam. A basic assumption is that the attenuation within each voxel is a fraction of the fluorescent intensity within that same voxel. A simulation program has been written where the parameters of the diffusion process within the microsphere can be controlled. Also the parameters for the attenuation calculation can be set, e.g. the assumed fraction of fluorescent intensity that act as attenuation. 3D datasets can be generated for visualization. Also intensity profiles can be generated along a diameter of the simulated sphere in the depth direction, since the intensity distribution is circularly symmetric in the lateral directions. Some comparisons are made to real microspheres, and the parameters are adjusted for closest resemblance. This adjustment can be done manually but an implementation using non-linear fitting of data is also presented. The simulated diffusion constant, fraction of intensity acting as attenuation, and the maximum intensity are fitted to experimental data from vertical slices of microspheres.

Standing-wave fluorescence microscopy, a method which utilizes interference to create a periodic excitation pattern along the optical axis, has been shown to provide improved axial resolution in thin, fluorescently labeled specimens. In each plane of focus, a complete standing wave data set is obtained by acquiring an image at each of three distinct positions of the interference fringes. Thicker specimens require through-focus data consisting of three images per plane. In this report we describe the recovery of information from this data using 3D image processing. The effective optical transfer function (OTF) of the standing wave microscope consists of the conventional OTF and two sidebands which are copies of the conventional OTF shifted axially by the spatial frequency of the interference fringes. The large gaps between the central band and the sidebands lead to significant ringing in the 3D reconstruction if linear deconvolution methods are employed. The use of non-linear, constrained image processing techniques has been shown to allow accurate extrapolation outside the OTF band limit. We demonstrate the extent to which the sidebands enhance recovery of information in the gaps, and provide a comparison between deconvolution using inverse-filtering and maximum-likelihood estimation.

Imaging in fluorescence microscopy is analyzed using a vectorial diffraction theory. Both conventional and confocal microscopy are considered. A fluorescent molecule is modeled as a radiating electric dipole. Two particular limiting cases are considered: the dipole can either freely rotate in space between excitation and emission, or is fixed in space. For each case, we average over all dipole orientations.

Living cells and other non-embedded biological specimens present a challenge in high- resolution microscopy because of their relatively low refractive index. It is well known that when oil-immersion objectives are used to look into these specimens, spherical aberration resulting from the index mismatch causes axial as well as transverse broadening of the 3D point spread function. Both effects reduce 3D resolution significantly, and cause the microscope to be axially shift-variant, thus complicating 3D deconvolution-based computational refinement. Axial shift variance should be much less a problem when water immersion objectives are used. With the aim of improving image data and our ability to refine it computationally, we have evaluated an apochromatic 40X/1.2 NA indirect water- immersion objective having an adjustable corrector for coverslip thickness. This work involves comparison of measured and computed 3D point spread functions, and comparison to the data acquired with an oil-immersion lens system. We show that focus-dependent spherical aberration is greatly reduced in the water-immersion system.

A novel technique that provides a functional response, similar to the point spread function, for the characterization of high NA lenses in real time is presented. This measurement technique employs absorption and subsequent fluorescence in a bulk solution of a suitable fluorophore. Both the dye and the solvent can be chosen to match the experimental conditions best for which the lens is tested, especially with respect to the refractive index of the solvent and the wavelengths of excitation and emission. The experimental results presented show the sensitivity of the new technique to different apodization conditions of the lens. Theoretically the experiments are modeled from the Rayleigh-Sommerfeld integral of scalar diffraction theory in the Kirchhoff approximation. Applying the shift-invariance approximation to save computing time, theoretical curves are calculated without any fitting parameters that show excellent agreement with the experimental results.

We investigate the effects of the apodization functions of a high-aperture objective on axial resolution in confocal bright-field microscopy. It is found that the measured apodization function for a commercial objective deviates significantly from the function predicted under the sine condition, which is usually assumed to be satisfied by a commercial objective. This result may explain the long-time discrepancy of the axial resolution between theoretical prediction and measured results, and also the observation of smoothly oscillating side peaks in the axial response when the objective is operated at an incorrect tube length. This finding is of importance when one tries to compensate for the spherical aberrations, with a certain accuracy, by the alteration of tube length of the objective. The deviation of apodization of a practical objective from the sine condition may be also responsible for the observation of the broadened axial response in confocal fluorescence microscopy including 4Pi confocal systems.

We describe the use of avalanche photodiodes as solid state photon counters in scanning confocal microscopy. Photon counters have been of limited use for moderate-to-rapid image acquisition speeds due to comparatively low saturation count rates. Several approaches offer the promise of real-time photon-limited image acquisition using off-the-shelf components and dark count rates which are fully adequate for imaging. We characterize in some detail a gated NPN switch/pulse-bias circuit which offers excellent, economical performance over a wide wavelength range, and discuss specific imaging applications.

In this paper we describe a new fiber confocal optical microscopy with miniature scanner and imaging optics that is being developed for in situ imaging applications. The microscope uses torsional mirrors to accomplish the scanning. The mirrors, which are 300 X 360 micrometers and 500 X 600 micrometers , respectively, are fabricated with silicon micromachining techniques and produce a raster scan at video rates suitable for real time imaging. The objective is an off-axis grating lens and has numerical aperture of 0.25. Imaging is monochromatic at (lambda) equals 0.6328 micrometers , and the system is capable of approximately 1.0 micrometers resolution over a 100 micrometers field of view. The cross sectional dimensions of the imaging head that contains the mirrors and the lens is 1.2 X 2.5 mm. We present the microscope architecture and discuss the design parameters, the limitations, and the performance tradeoffs that were faced in developing this microscope.

We describe a semiconductor laser confocal interference microscope that is capable of producing both confocal images and high resolution surface profiles. The system is based on an optical fiber interferometer together with injection current modulation. We also address the issue of resolution in phase imaging.

We describe a real time white light reflection confocal microscope incorporating an optical fiber bundle. We investigate the optical sectioning properties of the fiber bundle in detail.

We discuss the basis of color confocal and confocal spectroscopic imaging. We consider the question of chromatic aberration in detail.

We develop a theory for brightfield direct-view microscopy using finite-sized, multiple pinhole arrays in the source and detector planes. We show that the array geometry can be decoupled mathematically from the remainder of the optical system. In particular, we present an expression for the response to a point object and to plane reflector. While the theory applies to an arbitrary distribution of pinholes we consider in detail the case of square, hexagonal and interleaving Archimedean spiral arrays. We also consider the implications of the array configuration on both the optical sectioning strength and the light budget.

We have developed a prototype fluorescence microscope which, using tomographic image acquisition and reconstruction techniques, can automatically combine conventional and/or confocal image stacks taken at a number of orientations into a single, very-high-resolution 3D image. We use the term `microtomography' in a broad sense to denote digital image reconstruction from multiple imaging operations which are not necessarily projections. Our system holds a biological specimen inside a thin capillary tube which is rotatable over a 360 degree range beneath an immersion objective. 3D fluorescence image data volumes are acquired by frame-grabbing a through-focus series of 2D images at each angle of rotation. Digital reconstruction of the multi-angle data volumes produces a single very-high-resolution 3D image and involves algorithms which perform rotation, interpolation, alignment and normalization operations in frequency (Fourier) space.

We developed a double-axis fluorescence microscope as a new type of 3D microscope, in which two conventional microscopes are placed perpendicularly to one another. This optical microscope provides a solution to extend the spatial frequency band in the depth direction, which is greatly limited for conventional single-axis microscopes. In addition, this equipment has a capability of implementing a fast image acquisition system by simultaneously taking images through both microscopes. We evaluated the 3D imaging properties of our system using 3D observation images of a micro-fluorescent sphere. In order to reconstruct 3D images of micro-objects, we applied a CT-based method we proposed to the 3D microscopic imaging system. Our experiments using a biological micro-specimen verified usefulness and effectiveness of our system.

The Fourier transform Confocal Raman Microscope (FT-CRM) enables non-invasive 3D Raman spectroscopic analysis and visualization of chemically heterogeneous preparations. The instrument combines a confocal optical microscope with a visible light Fourier transform Raman spectrometer to acquire and analyze the Raman spectrum of light scattered from a voxel in the sample defined by the confocal optics. Scanning the sample relative to the confocal voxel and recording the Raman spectrum at each scan position generates a multi- dimensional data set encoding the spatially-varying compositional properties of the sample. We report here on the spatial and spectral FT-CRM image properties that includes recent work on correlation-based Raman spectroscopic imaging and application of parametric spectral estimators for robust Raman spectrum estimation.

We describe a Raman imaging microscope that produces high-fidelity, large format Raman images and Raman spectra at close to diffraction-limited spatial resolution. A silicon charge- coupled device is used as a high sensitivity array detector. Wavelength selection of Raman scattered emission is achieved by an acousto-optic tunable filter, which maintains image fidelity and provides either continuous or random wavelength selection. Laser illumination is delivered to the object by means of an infinity corrected microscope objective, either by a galvanometer scanning system or a widefield fiber optic. The laser scanning mechanism has higher power densities and provides Raman microprobe capabilities when stopped at a prescribed point. The fiber optic illumination scheme, however, is useful for delicate samples which might be damaged by the higher power densities generated by the point scanner mechanism and for sample alignment and system focusing. Instrument features, including factors that determine the system's spatial and spectral resolution, are discussed in detail. Images and spectra of test objects and samples that demonstrate the capabilities of this imaging spectrometer are presented. The potential of intrinsic chemical imaging is discussed in terms of its use in the analyses of a variety of chemical and biological samples.

A microscope having a high effective numerical aperture is achieved in an apparatus in which a real, 3D image is formed of an object placed along the axis of one of two facing concave mirrors and is acquired by a video camera, preferably a CCD, positioned along the axis of the other mirror. The image acquired by the CCD is electronically stored and then analyzed. Magnification is introduced by the spacing of the sensors of the CCD array. With exemplary CCD sensor spacing of 10 microns, resolution of about 10 microns (10^-5 meters) is achievable using commercial 8-inch diameter concave mirrors. The high numerical aperture of the reflecting microscope offers the advantages of being able to more rapidly acquire an image at a lower light level than is possible with a slit lamp microscope which requires that the object be physically scanned with a high intensity light.

Three-dimensional microscopy by computational deconvolution methods requires accurate knowledge of the point spread function (PSF) that characterizes the microscope. Experimental PSF's can only be measured over small regions about focus because the small objects necessary for PSF measurement are dim. Theoretical computation of the PSF requires accurate knowledge of all the experimental setup parameters. Some parameters may be difficult or impossible to measure. In blind deconvolution, the PSF and the specimen are estimated simultaneously, an under-determined problem with non-unique solutions. Most existing approaches to blind deconvolution rely on enforcing constraints on the specimen function and PSF, sometimes in ad-hoc ways. We derived a parametric blind deconvolution method by assuming that the PSF follows a mathematical expression with unknown parameters. The parameters are then estimated together with the specimen function. Preliminary results presented here show that this algorithm rapidly estimates the correct PSF.

The Maximum Likelihood based blind deconvolution (ML-blind) algorithm is used to deblur 3D microscope images. This approach was first introduced to the microscope community by us circa 1992. The basic advantage of a blind algorithm is that it simplifies the user interface protocols and reconstructs both the object and the Point Spread Function. In this paper we will discuss the recent improvements to the algorithm that robustize the performance and accelerate the speed of convergence. For instance, powerful and physically justified constraints are enforced on the reconstructed PSF at every iteration for robustization. A line search technique is added to the object reconstruction to accelerate the convergence of the object estimate. A simple modification to the algorithm enables adaptation for the transmitted light brightfield modality. Finally, we incorporate montaging in order to process large data fields.

We describe a prototype 3D optical microscopy system that utilizes parallel computing and high-speed networks to address a major obstacle to successful implementation of 3D dynamic microscopy: the huge computational demand of real-time dynamic 3D acquisition, reconstruction, and display, and the high-bandwidth demand of data transfer for remote processing and display. The system comprises image acquisition hardware and software, high- speed networks between acquisition and processing environments, parallel restoration using wavelet algorithms, and volume rendering and display in a virtual environment.

Maximum likelihood image restoration is a powerful method for 3D computational optical sectioning microscopy of extended objects. With punctate specimens, however, this method produces a few very bright isolated spots and dim detail around them is lost. The commonly used regularization methods (sieves and roughness penalty) decrease the amplitude of the bright spots, but do not avoid loosing dim detail. We derived an intensity regularization that decreases the amplitude of bright spots without loosing dim detail. In contrast with other regularization methods, this method does not increase significantly the computational complexity of the estimation algorithm.

The authors present a Hough transform based image segmentation algorithm for automated detection, counting, and measurement of particles in two- and three-dimensional microscopic digital image data sets. The algorithm has proven to be both sensitive and specific for the particles of interest, even in the presence of noise and blurring. We apply the algorithm to the problem of automated compilation of population statistics on the size and mass of zymogen granules in pancreatic acinar cells. We present results describing the performance of the algorithm on digital phantoms, and image data from conventional fluorescence and confocal microscopes.

Shape reconstruction from stereo images of fracture surfaces, which are obtained by scanning electron microscopy, is applied to research in micromechanics and fracture mechanics. The shape is reconstructed by matching the corresponding points on a pair of stereo images. To deal with the problem of false-target matching and to reduce computation time, the `interesting points', which usually are pixel locations of important features in each pair of images, are determined first, and the matching is only performed on these `interesting points'. Two algorithms are used for finding the `interesting points'. The first one selects the pixels that have the maximum value of local entropy. The second one selects pixels having the local maximum gray scale value after the image has been processed by a Laplacian of Gaussian filter. In order to reduce the effects of shadows, the logarithm of intensity is calculated before the image is filtered. The algorithm of matching corresponding points in the stereo images is based on the theory of cross correlation. To reduce the false-matching, a coarse and fine matching algorithm is developed. Finally, the height of the surface is calculated by applying stereo triangulation theory.

Three-dimensional reconstruction of biological specimens often requires the registration of successive sections of tissue. However, the introduction of fiducial markers is exceedingly difficult in some imaging modalities. We describe a fully automated registration method for use with transmission electron microscopy or other modalities that does not require the existence of fiducial markers. This system employs an error manipulation approach with progressive refinement that converges to the locally optimal registration across al considered objects. When used in conjunction with other methods for coarse refinement, a globally optimal solution is produced. The method can also be customized to the characteristics of the tissue involved. Among many of the possibilities, it can register based on contour location or, more interestingly, on contour shape, which effectively minimizes the surface tension of the resulting reconstruction. In addition, weighting factors are available to give increased consideration to larger objects when appropriate.

A framework for a computer-guided image acquisition procedure in confocal microscopy is described, which allows a 9-fold increase of the field of view and an extended imaging in axial direction by scanning aligned thick serial sections. Using an image compositing technique, complete respiratory units (acini) of small mammalian species (rat, mouse) could be investigated at high resolution. The quantification of the composed volume includes segmentation and topological investigation. With the help of parameters measured, a computer model of the acinus is designed to prepare functional simulations. In order to study flow of air in the acini, diffusion and the oxygen uptake, a set of mass transport equations is solved iteratively. The structural dynamics of ventilatory units during inhalation and expiration is also included.

Nomarski Differential-Interference-Contrast (DIC) microscopy is a widely used method for imaging transparent specimens that are not visible with ordinary light microscopy. DIC microscopy enhances contrast in the images of such specimens by converting differential phase changes to intensity variations via the method of light interference. These phase changes are introduced in light as it passes through regions of different refractive index within a specimen. In this paper, the development of an imaging model that describes 3D DIC imaging under partially-coherent illumination is presented. Our approach in deriving the model involves the derivation of a 2D model and its extension to three dimensions, assuming weak optical interactions within the specimen. The coherent limit of our 2D model coincides with existing DIC models. Model predictions generated with the coherent limit of the 3D model are compared to real DIC images acquired from imaging phantom specimens. It is shown that the model predictions resemble the real images obtained with the condenser aperture closed better than the images obtained with the aperture open. This result confirms the need for the general model that we have derived.

Optical coherence tomography (OCT) is an emerging alternative imaging tool to confocal microscopy. In layers beyond about 200 - 300 micrometers depth, an increasing fraction of multiple scattered photons begins to deteriorate diffraction limited axial and lateral resolution curves, which otherwise can be obtained only in very superficial layers (single-scatter regime). At greater depths, the contrast and resolution from OCT (and confocal microscopy) are determined by the parameters of the turbid medium rather than by the focusing optics. We have developed an analytical model to describe spatial resolution curves in homogeneous turbid media employing the heterodyne interferometric principle. Analogous to basic ideas from theoretical work done in the atmospheric LIDAR (Light Detecting and Ranging) community we derived the heterodyne detector signal from a mutual coherent function (MCF). The MCF is a function of the parameters of the focusing optics, the object position and the degree of coherence (lateral coherence length), which in turn characterizes the turbid medium. Axial resolution curves were acquired with our interferometer in reflection mode to characterize various turbid media by fitting the experimental data to simulation curves. Particularly when light propagates through suspensions of large particles before impinging on the object, a considerably loss in contrast (and even resolution) of the curves is noticeable. We studied the effects of a mirror and a diffuse plate serving as reflecting targets. In ex vivo tissue, we obtained a lateral coherence length on the order of 1 micrometers under the assumption of the validity of the model used.

Based on the principle of spatial autocorrelation of the focal field, improved resolution can be obtained over that obtainable in confocal microscopy. The method is presented and the imaging resolution is analyzed for two model objects. The method employs only one high NA lens for the, potentially robust, common path delivery of the autocorrelated excitation beams.

The aberrations introduced when focusing within a specimen with a refractive index equal to that of water using an oil-immersion objective are investigated. The peak intensity in the confocal point spread function drops by a factor of two for focusing less than 10 micrometers into the specimen. The effects of scaling of dimensions in the resulting images are discussed. The image exhibits an axial stretching by a factor of about 1.12.

Bases of constructive theory of formation in coherent light of the Fraunhofer diffraction (Fourier spectrum) patterns of the opaque 3D objects of constant thickness with flat internal surfaces are presented. Such theory is simple, physical obvious and at the same time sufficiently strict. It is based on the model of the equivalent diaphragms according to which the problem of light diffraction on volumetric bodies is reduced to the analysis of diffraction phenomena on the plane transparencies which are located in space. It permits to apply the standard Fourier-optical methods for the calculation in Kirchhoff-Fresnel approximation. This theory is developed and generalized for cases of formation and filtering the images and Fraunhofer diffraction patterns of the typical elements of extended bodies, including volumetric edge, 3D slit. Dependencies between the characteristic parameters of the diffraction patterns and geometrical dimensions of 3D slit are found on the basis of the behavior of the Fourier spectra of extended objects. Peculiarities of coherent optical processing of 3D objects are investigated in detail by the example of high-frequency filtering (contouring) of volumetric edge with absorbing, reflecting and grey inner surfaces.

The resolution of standard 3D optical microscopies--both confocal and widefield+deconvolution--is substantially worse in the axial direction than in the lateral plane. This weakness is caused by the fact that the objective lens in a conventional microscopy can only access light emitted within a limited angle of the optical axis. We are developing several new widefield techniques in which light is collected over a larger set of angles by using two separate, opposing objective lenses to observe and/or illuminate the sample from both sides simultaneously. This gives access to substantially more spatial information about the sample than can normally be reached. This information is in the form of relative phase correlations between the two image beams, and thus cannot be accessed by studying either beam alone, but can be detected through the interference of the two image beams if they are combined coherently on a single image detector. The added information allows computational 3D reconstruction to be made with drastically improved axial resolution compared to conventional widefield or confocal microscopies. Experimentally measured transfer functions from our prototype microscope show excellent qualitative agreement with theoretical expectations, and 3D reconstructions of test samples confirm the expected improvement in axial resolution.

The laser confocal microscope (LCM) is now an established research tool in biology and materials science. In biological applications, it is usually employed to detect the location of fluorescent market molecules and, under these conditions, signal levels from bright areas are often < 20 photons/pixel (from the specimen, assuming a standard 512 X 768, 1 sec. scan). Although this data rate limits the speed at which information can be derived from the specimen, saturation of the fluorophor, photobleaching of the dye, and phototoxicity prevent it being increased. Currently, most LCMs use photomultiplier tubes (PMT, QE equals 1 - 30% 400 - 900 nm). By contrast, rear-illuminated, scientific charge-coupled devices (CCD) now routinely readout the signal from square sensors approximately 30 micrometers on a side with a QE of 80 - 90%, a noise of only +/- 3 e^-/pix and with no multiplicative noise. For this reason, in 1989, one of us (JJ) developed a rear-illuminated, single-channel Si sensor, called the Turbodiode, employing some of the sophisticated readout techniques used to measure charge in a scientific CCD. We are now extending this work to a device in which a single 36 X 36 micrometers sensor is read out through a low-noise FET charge amplifier with a reset circuit and then passed to a correlated, double-sampling digitizer. To maintain the desired +/- 3 e noise level at the relatively high data rate of 1 MHz, our new device utilizes 64 separate readout amplifier/digitizer systems, operating in sequence. The resulting detector is more compact, efficient and reliable than the PMT it replaces but as its sensitive area is smaller than that of a PMT, it will require auxiliary optics when used with any LCM having a large (mm) pinhole. As the signal light is parallel, a simple lens mounted axially and with the CCDiode at its focus would suffice. Future versions may use 3 X 3 or 5 X 5 arrays of sensors to `track' the confocal spot as it is deflected by inhomogeneities of the specimen, change its effective size or shape or detect system misalignment.

This brief proceedings paper presents an introduction to our adaptation of the principles of phase shifting interferometry to a laser feedback interferometer. The application of these methods allows a direct measurement of both the optical path length and the fringe modulation. Examination of the spatial variation of both of these quantities over an object's surface provides a quantitative map of the geometry of a sample's surface. We demonstrate that discrete phase shifting methods can be used to accurately measure optical path length changes and fringe modulation.