Confocal laser scanning microscopy is used in many different fields of research nowadays. Therefore, high spatial resolution is required as well as high temporal resolution. Further, the quality of the resulting images has to be considered. Doing imaging with laser scanning microscopes, the balance between spatial resolution, speed and signal-to-noise ratio has to be defined for every specimen and experiment individually. Special adaptations to standard laser scanning microscopes improve the efficiency of the Leica TCS SP2 and Leica TCS SP2 RS for high-resolution low-noise imaging. Here, we want to report from the acousto-optical beam splitter, the resonant K-scanner and new objectives.

Recent developement in confocal microscope systems allow to tune the spectral characteristics of those elements which separate excitation and emission. The new devices replace the classical filter-devices, which have been used in fluorescence microscopy for many years. Excitation filters have been replaced by acousto optical filters, which work as multichannel dimmer devices and allow fine tuning of excitation energy as well as tuning the excitation ratio of multiple excitations. Emission filters have been replaced by tunable multiband detectors, which allow tuning of emission band width for higher efficiency and better band separation. The multiband detector also allows spectral scanning modes. Dichroitic beam splitters have been replaced by acousto optical beam splitters (AOBS) giving significantly increased signal efficiency and allow fast switching of excitation modes without moving mechanical parts. All three devices together lead to much better primary images, either giving highly improved signal to noise ratio or much less photobleaching (which is identically with viability of life samples).

The abundance of naturally fluorescing components (autofluorophors) encountered in environmentally sourced samples can greatly hinder the detection and identification of fluorescently labeled target using fluorescence microscopy. Time-resolved fluorescence microscopy (TRFM) is a technique that reduces the effects of autofluorescence through precisely controlled time delays. Lanthanide chelates have fluorescence lifetimes many orders of magnitude greater than typical autofluorophors, and persist in their luminescence long after autofluorescence has ceased. An intense short pulse of (UV) light is used to excite fluorescence in the sample and after a short delay period the longer persisting fluorescence from the chelate is captured with an image-intensified CCD camera. The choice of pulsed excitation source for TRFM has a large impact on the price and performance of the instrument. A flashlamp with a short pulse duration was selected for our instrument because of the high spectral energy in the UV region and short pulse length. However, flash output decays with an approximate lifetime of 18μs and the TRFM requires a long-lived chelate to ensure probe fluorescence is still visible after decay of the flash plasma. We synthesized a recently reported fluorescent chelate (BHHCT) and conjugated it to a monoclonal antibody directed against the water-borne parasite Giardia lamblia. Fluorescence lifetime of the construct was determined to be 339μs ± 14μs and provided a 45-fold enhancement of labeled Giardia over background using a gate delay of 100μs. Despite the sub-optimal decay characteristics of the light pulse, flashlamps have many advantages compared to optical chopper wheels and modulated lasers. Their low cost, lack of vibration, ease of interface and small footprint are important factors to consider in TRFM design.

We have developed a new prototype of a confocal scanning laser ophthalmoscope that incorporate relatively low-cost adaptive optics to correct for wavefront aberrations induced in the exit path of the eye and the optical setup components. The scanning part of the system consists of two galvanometric scanners, and the adaptive optics part contains a membrane deformable mirror in conjunction with a Hartmann-Shack wavefront sensor. The system allows to register images of the retina with infrared illumination at a 15 Hz frame rate and with a variable viewing angle in the range of 1° to 10°. We show first results obtained with the system with images of a test target in an artificial eye and with imaging of the living human retina. We compare images obtained without and with the adaptive optics part activated. In preliminary images, retinal features down to a size of ~25 μm have been resolved with the application of adaptive optics.

We have developed a novel confocal microscopy system, which can perform both axial and lateral scanning at an ultra high speed. A rotary mirror array (RMA) consists of 36 small mirrors deployed on the surface of a round plate with a small tilting angle and a discrete rotational symmetry. When the RMA rotates at a constant speed, the focal point is scanned inside a sample periodically and linearly. We have demonstrated a scanning rate of 2400 lines per second, which can be readily improved to more than 20000 lines per second. Our scheme can also be used in a fast scanning multi-photon microscope.

Real-time optically sectioned microscopy is demonstrated using an AC-sensitive detection concept realized with smart CMOS image sensor and structured light illumination by a continuously moving periodic pattern. We describe two different detection systems based on CMOS image sensors for the detection and on-chip processing of the sectioned images in real time. A region-of-interest is sampled at high frame rate. The demodulated signal delivered by the detector corresponds to the depth discriminated image of the sample. The measured FWHM of the axial response depends on the spatial frequency of the projected grid illumination and is in the μm-range. The effect of using broadband incoherent illumination is discussed. The performance of these systems is demonstrated by imaging technical as well as biological samples.

We present a new imaging modality using nanometric beads as multiple local probes. The positions of the beads are determined in three dimensions using white-light interference microscopy, by over-sampling and fitting the images. We measured the deformation of a Laponite gel with 100 nm resolution.

The Quadrature Tomographic Microscope measures the amplitude and phase of an image. This information allows the user to see contrast features not available in other microscopes, and is critical to any three-dimensional reconstruction. We report on development and use of test objects to measure the accuracy and repeatability of phase measurements. A simple binary phase grating, a series of glass beads, and preimplantation mouse embryos were used in these experiments. The gratings were fabricated on high-quality fused-silica substrates whose transmission phase error was determined to be less than one-tenth wave error across their 25 mm diameter before fabrication. The phase step of the binary phase grating was measured using both the optical quadrature technique and the usual fringe-counting techniques applied to the raw data. Phase unwrapping techniques were validated by measuring the diameter of glass beads of a known size. Results are presented showing that the phase measurements agree with each other, with the known data, and with the spatial resolution in preimplantation mouse embryos. More complicated objects will be fabricated in the future to validate 3-D imaging techniques.

Images of the microspheres are studied in three-dimensions by using the confocal and conventional scanning polarization microscopes. It is found that the polarization of the detected signals is mainly parallel to the initial polarization which is due to the high extinction coefficient of the confocal system. Arc pairs are observed at the edge of the microspheres with the conventional polarization microscope with a crossed analyzer. Theoretical analysis are given by using the vector field theory and the image formations of the two systems.

The focused radially polarized Cylindrical Vector (CV) beam has been found to have a strong longitudinal field component at focus in the direction of propagation of the beam. It has been shown that this field component can be used to excite a second harmonic (SH) signal at a surface interface. Th spatial intensity profile that excites the SH signal has ben predicted to have a smaller lateral extent than a similarly focused linearly polarized beam. The possibility to use this longitudinal field component for SH surface imaging is investigated.

This paper describes a multi-view stereoscopic 3-D display. The display technology does not involve goggles, polarized glasses, colored glasses or any other eyewear. It allows parallax, in that as the viewer moves their head left and right, the viewer can 'see around corners.' The method involves a microlens array on top of a liquid crystal display. The microlens is designed to project multiple views to multiple eye positions. The distance between stereo eye positions is the average distance between eyes of 55 mm. As many as eight views are interlaced on the display and fanned out by the microlens array. Thus 4 stereo pairs can be observed, each pair from a unique angular perspective. For this system, since multiple views are available, seeing the 3-D effect is much easier. In addition, the 3-D effect can be seen far off axis; so more than 1 person can view the display at the same time.

Current high-performance computers and advanced image processing capabilities have made the application of three dimensional visualization objects in biomedical images facilitate the researches on biomedical engineering greatly. Trying to cooperate with the update technology using Internet, where 3-D data are typically stored and processed on powerful servers accessible by using TCP/IP, we held the results of the isosurface be applied in medical visualization generally. So in this system we use the 3-D file format VRML2.0, which is used through the Web interface for manipulating 3-D models. In this program we implemented to generate and modify triangular isosurface meshes by marching cubes algorithm, using OpenGL and MFC techniques to render the isosurface and manipulate voxel data. This software is more adequate visualization of volumetric data. The drawbacks are that 3-D image processing on personal computers is rather slow and the set of tools for 3-D visualization is limited. However, these limitations have not affected the applicability of this platform for all the tasks needed in elementary experiments in laboratory or data preprocessed. With the help of OCT and MPE scanning image system, applying these techniques to the visualization of rabbit brain, constructing data sets of hierarchical subdivisions of the cerebral information, we can establish a virtual environment on the World Wide Web for the rabbit brain research from its gross anatomy to its tissue and cellular levels of detail, providng graphical modeling and information management of both the outer and the inner space of the rabbit brain.

The analysis of cell and pathogen movement and motility is a major topic in cell biology for which computerized methods are most needed. This study proposes a method to detect and track multiple moving biological objects in image sequences acquired through fluorescence video microscopy. The method enables the analysis of video microscopy image sequences in order to obtain reliable quantitative data such as number, position, speed and movement phases. The method consists of three stages. A stage of detection is performed through a multi-scale analysis of images using an undecimated wavelet transform. The next stage is the prediction of the state of each detected spot in the next frame using a Kalman filter and an adapted model. Then comes a stage of data association which constructs the tracks and refines the filters. Once all moving objects have been assigned with unique spatio-temporal paths, trajectories are analyzed in terms of different parameters relevant to the motility analysis of biological objects.

We present results from 2D Fourier analysis on 3D stacks of images obtained by confocal laser scanning reflectance microscopy (CLSM) and two-photon fluorescence microscopy (2PM) on human skin in vivo. CLSM images were obtained with a modified commercial system (Vivascope1000, Lucid Inc, excitation wavelength 830 nm) equipped with a piezo-focusing element (350 μm range) for depth positioning of the objective lens. 2PM was performed with a specially designed set-up with excitation wavelength 730 nm. Mean cell size in the epidermal layer and structural orientation in the dermal layer have been determined as a function of depth by 2D Fourier analysis. Fourier analysis on microscopic images enables automatic non-invasive quantitative structural analysis (mean cell size and orientation) of living human skin.

Implicit active contour method are a powerful technique for segmentation and tracking of mobile objects such as biological cells observed in videomicroscopy. However, the lack of control on the topology changes in this approach often leads to undesirable contour fusions when previously distinct objects enter in close contact. To overcome this limitation, we propose to modulate the current image by a "ridge" which discourages contour motion towards neighboring objects, thus inhibiting contour fusions. We show applications of this method on both synthetic images and real images from cellular imaging.

In this paper we present a scheme for real time segmentation of histological structures in microscopic images of normal and neoplastic mammary gland sections. Paraffin embedded or frozen tissue blocks are sliced, and sections are stained with hematoxylin and eosin (H&E). The sections are then imaged using conventional bright field microscopy. The background of the images is corrected by arithmetic manipulation using a "phantom." Then we use the fast marching method with a speed function that depends on the brightness gradient of the image to obtain a preliminary approximation to the boundaries of the structures of interest within a region of interest (ROI) of the entire section manually selected by the user. We use the result of the fast marching method as the initial condition for the level set motion equation. We run this last method for a few steps and obtain the final result of the segmentation. These results can be connected from section to section to build a three-dimensional reconstruction of the entire tissue block that we are studying.

An approximate model for optical-sectioning microscopy describing depth-varying imaging is developed. The model incorporates changes in the point-spread function due to refractive index mismatch between the immersion medium and the specimen, which causes spherical aberration that worsens with increasing depth under the coverslip. Comparison of model predictions to measured images from a bead phantom shows that the approximate model captures the main features in the data. The model presented in this paper is the first step towards depth-variant image estimation for optical-sectioning microscopy. An expectation maximization algorithm for maximum-likelihood restoration based on this model is also presented.

Recognizing spatial relationships of neuro-degeneration in brains exposed to organic solvents is difficult when working from 2-dimensional serial slices. Recent advances in software have allowed the assembly of serial sections of stained tissue into a 3-dimensional (3D) representations. Appropriately chosen stains indicative of specific pathology can be highlighted and the 3D representation of its spatial distribution within the organ displayed. The purpose of this work was to develop a method for visualizing the spatial distribution of neuronal degeneration following organic solvent exposure. A cupric silver stain highly specific for degenerating neurons was used to identify neuronal degeneration in 83 serial histologic sections of brains of rodents exposed to toluene. Brain sections were scanned at 600 dpi using a grey-scale protocol. Scans were assembled into 3D images which were further processed into stereo pairs. Grey-scale scans were compared to the original sections in order to establish grey-scale ranges for healthy and damaged tissue and artifact staining. The respective categories then were assigned pseudo-colors to improve contrast.

We describe a simple modification to a rigid endoscope so as to provide both high quality conventional endoscopic as well as and confocal endoscopic images of reasonably accessible regions of the body in real time. The systems are based around either host lenslet-array tandem scanning microscope together with laser illumination or a structured illumination approach together with a conventional incoherent illumination source. Images taken in fluorescence are presented using this combined conventional and confocal endoscope.

Using both vibrational sum-frequency generation and fluorescence microscopy, the phase behavior of a DPPC lipid monolayer on water is investigated as a function of surface pressure. The vibrational specificity of the sum-frequency generation techniques permits determining the order of the alkyl chains, as well as the average orientation of the terminal methyl group. A novel -- and extremely sharp -- phase transition is observed at low compression, which is attributed to a curling of the alkyl chains due to increased exposure of the (hydrophobic) alkyl chains to the water surface.