Optical coherence tomography (OCT) is an emerging, high-resolution near-infrared imaging and microscopy technique. The axial and transverse resolutions in OCT can each be analyzed independently, with the axial resolution inversely proportional to the spectral bandwidth of the optical source and the transverse resolution defined by standard Gaussian beam optics. While high numerical-aperture objectives are preferred to improve the transverse resolution, the reduced confocal parameter limits the depth-ranging capabilities of OCT and more complex en face imaging with focus tracking must be employed. We present a method for increasing the apparent transverse resolution in OCT outside of the confocal parameter using Gaussian beam deconvolution of adjacent axial scans, and thereby reducing the limitations associated with the hourglass profile of a tightly focused Gaussian beam. Specifically, our method determines how measurements depend on the object when blurred with a Gaussian beam, and subsequently finds an estimate of the original object. Possible reconstruction estimates are explored and evaluated using a variety of regularization techniques as well as estimation maximization algorithms. Numerical simulations demonstrate effectiveness of each technique. When applied to experimentally-acquired OCT data, the use of these algorithms can improve the apparent transverse resolution outside of the confocal parameter, extending the comparable confocal parameter range along the axial direction. These results are likely to further improve the high-resolution cross-sectional imaging capabilities of OCT.

Light that would typically be discarded at a confocal microscope's detector pinhole will be collected and processed to allow a reduced spatial sampling rate and thus an improved image acquisition time. It is shown that collecting and appropriately processing the out-of-focus light will allow an axial sampling rate below that specified by the Nyquist criterion. To achieve this, a central detector pinhole and a number of out-of-focus regions are collected concurrently. This corresponds to imaging through several different channels, with differing point spread functions, in parallel. Since the spatial sampling rate is below the Nyquist frequency, aliasing occurs in the data from each of the channels. However, since the point spread functions are different, the aliasing effects are different in each channel. This allows the ensemble of aliased images to be processed into a single dealiased and deconvolved image. This potential utility of out-of-focus light is demonstrated through simulated examples for differing collection schemes and scanning rates. Results are shown for under-sampling by up to a factor of four. Collecting the out-of-focus light also improves instrument collection efficiency.

We present our recent results on the development of three-dimensional (3-D) optical computed- tomography microscope. The instrument is a novel imaging device for the 3-D visualization and quantitative analysis of absorption-stained biological samples. The first instrument developed by our group at the BC Cancer Research Centre used a digital micromirror device (DMD) as a spatial light modulator to control the angles of illumination. This new embodiment employs an optical scanner instead of the DMD. The optical scanner is placed in the illumination path of the microscope system, conjugate to the field plane. The optical system includes also two high numerical aperture objective lenses, a sample stage, a light source, and a CCD camera. Projections are acquired by illuminating a specimen at a number of selected angles within the numerical aperture of the objective (0 < φ < 135°). A new reconstruction algorithm that employs both transform-based and iterative methods is developed to address the limited-angle reconstruction problem. A transform-based reconstruction is used as an initial starting point for the following iterative reconstruction. A feedback correction of the reconstructed image is made on each iteration step. The algorithm enables to incorporate previously known information about the object into the reconstruction process, and improves the reconstruction accuracy. Microscopic 3-D volume reconstructions of quantitatively absorption-stained cells have been generated. The system enables one to look at multiple optical levels of a specimen, and at more natural tissue architecture, including intact cells. Axial and lateral resolutions were measured to be better than 6 microns.

When recording three-dimensional (3D) images by the method of optical sectioning microscopy, each optical section contains the in-focus information plus out-of-focus contributions that obscure the in-focus detail and reduce contrast. There are several methods to remove or prevent the out-of-focus contributions from the stack of optical sections. One such method is image estimation -the use of a computer program based on a mathematical description of the microscope to remove the out-of-focus contributions. Another method is the use of structured illumination and a simple arithmetic operation to obtain a image that in which the out-of-focus contributions are greatly reduced. We derived a method for image estimation that uses the images collected from the structured-illumination microscope. The method improves the resolution of small detail over that possible with the structured illumination using the simple arithmetic formula.

The fluorescence confocal microendoscope provides high-resolution, in-vivo imaging of cellular pathology during optical biopsy. There are indications that the examination of human ovaries with this instrument has diagnostic implications for the early detection of ovarian cancer. The purpose of this study was to develop a computer-aided system to facilitate the identification of ovarian cancer from digital images captured with the confocal microendoscope system. To achieve this goal, we modeled the cellular-level structure present in these images as texture and extracted features based on first-order statistics, spatial gray-level dependence matrices, and spatial-frequency content. Selection of the best features for classification was performed using traditional feature selection techniques including stepwise discriminant analysis, forward sequential search, a non-parametric method, principal component analysis, and a heuristic technique that combines the results of these methods. The best set of features selected was used for classification, and performance of various machine classifiers was compared by analyzing the areas under their receiver operating characteristic curves. The results show that it is possible to automatically identify patients with ovarian cancer based on texture features extracted from confocal microendoscope images and that the machine performance is superior to that of the human observer.

A recently developed algorithm is applied to calculate a state space realization of a 3D microscopy image set. It is based on interpreting the image set as the impulse response of a 3D separable system. As an application it is shown how this algorithm, combined with approximation steps, can be used to suppress noise in 3D experimental point spread functions. The approach was motivated by a well known problem that a noisy point spread function degrades the results of deconvolution algorithms for the restoration of 3D fluorescence microscopy image sets. The proposed approach can also be applied to 3D fluorescence microscopy image sets of cells.

Despite advances in the measurement of lymphatic flow, little is known about the actual velocities of flow in microlymphatic (~100 um diameter) vessels. In this paper, video microscopy and particle tracking methods were adapted and integrated with an ultra high-speed imaging camera to obtain measurements of high-speed lymph velocities that previous systems were incapable of measuring. In this study, a mesenteric microlymphatic vessel in a loop of the small intestine of a male Sprague-Dawley rat was exteriorized and imaged at a rate of 500 frames per second (fps) for several contraction sequences. Lymph velocity was shown to fluctuate cyclically with the vessel wall contractions and ranged from -1 to 4 mm/sec through a ten second sequence.

In confocal microscopy, the laser excitation must be near the fluorochrome absorption peak to efficiently excite the fluorochrome. However, most fluorochromes emitting in the red to infrared have absorption and emission spectra with overlapping wavelength ranges, i.e. small Stokes shifts. As a result, the laser excitation extends into the fluorescence wavelengths hindering separation of the reflected laser signal from the information-containing fluorescence signal. Therefore, compromises are necessary: (1) the entire reflected laser spectrum is filtered, eliminating some of the fluorescence signal; or (2) the entire fluorescence signal is recorded, including unwanted reflected laser light, increasing the background noise. These compromises must be addressed, even with fluorochromes - like Cy5 - having a relatively large Stokes shift. Cryogenically cooled diode lasers eliminate the need for these compromises by tuning the output wavelength away from the fluorochrome emission. By separating the excitation and emission spectra, the confocal fluorescence signal-to-noise ratio can be increased by filtering more of the reflected laser emission, without losing valuable fluorescence information. However, this results in slightly less efficient excitation of the fluorochrome. We will present spectrophotometric analyses of fluorochrome absorption, fluorescence, and diode laser emission as a function of diode operating temperature. We will show that as the diode lasers are cooled their output power increases and more than compensates for the lower fluorochrome excitation, resulting in significantly more intense fluorescence. Thus, by tuning the diode laser, more fluorescence information and less reflected laser light reach the detector, creating images with greater intensity and less background noise.

A multi-spectral confocal microendoscope (MCME) for in-vivo imaging has been developed. The MCME employs a flexible fiber-optic catheter coupled to a slit-scan confocal microscope with an imaging spectrometer. The catheter consists of a fiber-optic imaging bundle linked to a miniature objective and focus assembly. The focus mechanism allows for imaging to a maximum tissue depth of 200 microns. The 3mm diameter catheter may be used on its own or routed though the instrument channel of a commercial endoscope. The confocal nature of the system provides optical sectioning with 3 micron lateral resolution and 30 micron axial resolution. The system incorporates two laser sources and is therefore capable of simultaneous acquisition of spectra from multiple dyes using dual excitation. The prism based multi-spectral detection assembly is typically configured to collect 30 spectral samples over the visible range. The spectral sampling rate varies from 4nm/pixel at 490nm to 8nm/pixel at 660nm and the minimum resolvable wavelength difference varies from 8nm to 16nm over the same spectral range. Each of these characteristics are primarily dictated by the dispersion characteristics of the prism. The MCME is designed to examine cellular structures during optical biopsy and to exploit the diagnostic information contained within the spectral domain. The primary applications for the system include diagnosis of disease in the gastro-intestinal tract and female reproductive system. In-vitro, and ex-vivo multi-spectral results are presented.

We demonstrate a simple and light-efficient way of generating non-diffracting Bessel beams for use in confocal microscopy. A number of imaging modalities using such beams is discussed. Preliminary experimental results including brightfield, fluorescence and two-photon images are presented.

Microlenses have been implemented in confocal systems successfully as components of aperture arrays and as arrays of objective lenses. The use of the novel technology of variable focal length (VFL) microlenses in the confocal system is investigated. The use of VFL microlenses as an aperture array in conjunction with an optical fiber as a pinhole array is examined. Axial responses of the system where measured and the Full-Width Half Maximum (FWHM) found to be ~16μm. VFL microlenses as an array of objective lenses is investigated with a novel method for scanning in the axial direction presented. By variation of the focal length of the lenses the focal plane can be scanned through the sample without the mechanical movement of the sample or the objective lens, we have named this 'focal scanning'. It is shown that the limiting factor with this type of scanning is the low numerical aperture (NA) of the microlenses available. Both focal scanning and conventional scanning are examined for this experimental set-up.

We present a microscope set-up designed to produce three-dimensional images of the internal structures of various samples with high spatial resolution (a few tens of nanometers in axial and transverse directions). This level of resolution is reached by the use of nanometric sub-wavelength spherical gold beads as multiple local probes, dispersed in the hollow structures. The exploration, by Brownian motion, of the internal structures allows their three-dimensional reconstruction.

A confocal reflectance theta line-scanner is being developed for imaging human tissues in vivo. The theta line scanner design potentially offers a newer alternative to current point scanners that may simplify the optics, electronics and mechanics and lead to smaller, inexpensive confocal microscopes. An oscillating galvanometric mirror directly scans in the pupil of a cylindrical lens and one-half of an objective lens, to produce a focused, scanned line in the object plane within tissue. Backscattered light is collected by the other half of the objective lens and focused onto a linear CMOS detector. The illumination is with a diode laser at 830 nm and imaging with a 10X, 0.8 NA water immersion lens. The illumination and detection paths are thus oriented at an angle (theta) to each other, and are separate everywhere except in the confocal plane. This configuration eliminates back-scattered light from optical components and enhances contrast. Optical design analysis has been verified with experimental results, demonstrating lateral resolution on the order of 1 um and optical sectioning (axial resolution) better than 5 um within living human skin. A Fourier optics-based analytical model is in progress to evaluate line spread functions versus illumination and detection pupil conditions. Nuclear and cellular detail is imaged in the epidermis of human skin in vivo and ex vivo (freshly excised specimens). Such a scanner may prove useful for imaging human tissues in clinical and intra-operative settings.

Real time in vivo optical coherence tomography (OCT) imaging of the adult fruit fly Drosophila melanogaster heart using a newly designed OCT microscope allows accurate assessment of cardiac anatomy and function. D. melanogaster has been used extensively in genetic research for over a century, but in vivo evaluation of the heart has been limited by available imaging technology. The ability to assess phenotypic changes with micrometer-scale resolution noninvasively in genetic models such as D. melanogaster is needed in the advancing fields of developmental biology and genetics. We have developed a dedicated small animal OCT imaging system incorporating a state-of-the-art, real time OCT scanner integrated into a standard stereo zoom microscope which allows for simultaneous OCT and video imaging. System capabilities include A-scan, B-scan, and M-scan imaging as well as automated 3D volumetric acquisition and visualization. Transverse and sagittal B-mode scans of the four chambered D. melanogaster heart have been obtained with the OCT microscope and are consistent with detailed anatomical studies from the literature. Further analysis by M-mode scanning is currently under way to assess cardiac function as a function of age and sex by determination of shortening fraction and ejection fraction. These studies create control cardiac data on the wild type D. melanogaster, allowing subsequent evaluation of phenotypic cardiac changes in this model after regulated genetic mutation.

We have developed a white-light interference microscope as an alternative technique to conventional optical coherence tomography (OCT). The experimental setup is based on a Linnik interferometer illuminated with a tungsten halogen lamp. En face tomographic images are obtained in real-time without scanning by computing the difference of two phase-opposed interferometric images recorded by a CCD camera. The short coherence length of the source yields an optical sectioning ability with 0.7 μm resolution (in water). Transverse resolution of 0.9 μm is achieved by using high numerical aperture microscope objectives. A shot-noise limited detection sensitivity of 90 dB can be reached with ~ 1 s acquisition time. High-resolution images of mouse and tadpole embryos are shown.

Dynamic imaging of cardiomyocytes was performed with a simultaneous detection second harmonic generation (SHG), third harmonic generation (THG) and multiphoton excitation fluorescence (MPF) microscope. The fast scanning system of ~12 frames/second synchronized with multichannel detection provided the possibility of imaging three dimensional static and two dimensional dynamic structures of cardiomyocytes. The SHG images highlighted the myofibrils of the cardiomyocytes while THG images revealed the locations of mitochondria. Dynamic data showed that during imaging, chaotic nanocontractions took place inside the cardiomyocytes. The time series of THG images reveled large intensity fluctuations "flickering" in the regions of mitochondria. The flickering in THG correlated with the flickering in MPF. Addition of the uncoupler FCCP inhibited flickering in THG and MPF, and also inhibited nanocontractions. The simultaneous imaging with SHG, THG and MPF proved to be a very powerful microscopy tool for investigation of interactions of different organelles inside a cell.

A mulitmodal microscope provides a researcher with the ability to capture images with multiple sources of contrast. Previous attempts at multimodal microscopy have been limited to capturing images at different instruments and digitally registering the images empirically using features common in the specimen. Other multimodal microscopes have combined different microscopy techniques, but have been limited in their sources of contrast. We present a unique microscope which containes 5 different imaging modalities, with four different sources of contrast. Quadrature microscopy measures the phase of the electric field of the light transmitted through an optically transparent specimen, by interference with a reference beam, using a polarimetric technique to resolve the ambiguities between phase and amplitude. Differential Interference Contrast (DIC) measures the phase gradient in an optically transparent specimen by combining two spatially separated images interferometrically. Confocal reflectance measures index of refraction changes in a plane or reflection from small scatterers. Confocal fluorescence and 2-photon laser scanning microscopy, measure fluorescent signatures of a specimen. The last three of these are inherently capable of producing three-dimensional images directly, through localized probing.

Automated quantitative (i.e. stochiometric) analysis of tissues is of eminent importance in the understanding of all interactions between cells in their natural environment. In tissue cytometry a solid trigger is necessary in order to unequivocally differentiate between cellular and non-cellular events. This can be best performed by nuclear staining. Aim of this study was to analyze a brain tissue section by laser scanning cytometry (LSC) in order to depict the threedimensional distribution of nuclei in the tissue. To this end the section was measured in several foci and different nuclei detected in several depths of the tissue were assigned to the respective layer. Frozen sections of formalin-fixed rat or human brain tissue (120μm thickness) were incubated with propidiumiodide (PI) (50μg/ml) and covered on slides. For analysis by the LSC propidiumiodide was used as trigger. After a first analysis focussed on the top of the tissue, the focus was adjusted in 30μm steps deeper into the tissue. Per analysis data of at least 50,000 cells were acquired. After finishing measurements from all depths of the field were merged, i.e. data were combined into a composite data file. With the special features of the LSC it was possible to develop a method depicting the threedimensional distribution of the nuclei in solid tissue sections. LSC can be useful tool for this relatively new field of solid tissue cytometry termed tissomics. After evaluation of methods like this, so far not available data can be analysed for diagnostic purposes. By these studies we intend to demonstrate the power of the LSC for the routine pathological use. This should add up to the bright versatility of applications for the LSC as a cytometric instrument suitable for high throughput and high content analysis.

Lateral resolution enhancement in confocal self-interference microscopy (CSIM) is evaluated. CSIM, which uses the birefringence of the calcite plate to generate self-interference pattern, sharpens the central lobe of the effective spot. Numerical simulation results of two-dimensional imaging performances are presented. Two-point resolution of 149nm is achieved, which is enhanced by nearly 100% compared to that of confocal microscopy.

Using an acousto-optic beam deflector and a line scan camera, we constructed a non-mechanical, slit-scanning confocal microscope. It generates two-dimensional 512x512 images with speed of 60 frames/sec, which can be easily expanded, to 100 frames/sec. The measured axial and lateral resolutions of the system are 3.3 mm and <1 mm, respectively, with a 50X objective. This simple design can produce frames rates faster than other commercial products with comparable resolution, which may be useful for analysis of rapid interactions in various biological applications.

Three-dimensional (3D) microscopy is a huge field that includes several microscopy techniques. Among these techniques the confocal microscopy is widely known and used due to its improved axial resolution or optical sectioning ability. However its high cost and image acquisition time as well as low signal-to-noise ratio are important drawbacks. Other techniques use wide-field structured illumination to get depth information from the sample. The detectors in commercial microscopes are normally of the point detector type (PMTs) in confocal and area sensors (CCDs) in wide-field microscopy. Examples of microscopes that take advantage of lower cost and faster readout cycles of linear image sensors are very uncommon. The purpose of this work is to develop a low-cost microscopy technique for obtaining 3D images with lower acquisition time and higher signal-to-noise ratio (SNR) than in classical confocal microcopy. It uses a linear CMOS image sensor and specific reconstruction algorithms will be able to extract 3D information from the distribution of light intensity collected. The sensor readout circuitry has been developed and tests are currently running in a laboratory prototype in reflection mode with an epi-illuminated configuration of scanning-stage type. The light source is a commercial incandescent lamp with regulated intensity. For the development, test and optimization of the reconstruction algorithms the object is mounted on a 3-axis translation stage to scan the object in three perpendicular directions. Contrast and resolution results obtained using a resolution test slide and preliminary images of integrated circuit (IC) bonding are presented.

Broadband interferometry is an attractive technique for the detection of cellular motions because it provides depth-resolved interferometric phase information via coherence gating. Here a phase sensitive technique called spectral domain phase microscopy (SDPM) is presented. SDPM is a functional extension of spectral domain optical coherence tomography that allows for the detection of cellular motions and dynamics with nanometer-scale sensitivity. This sensitivity is made possible by the inherent phase stability of spectral domain OCT combined with common-path interferometry. The theory that underlies this technique is presented, the sensitivity of the technique is demonstrated by the measurement of the thermal expansion coefficient of borosilicate glass, and the response of an Amoeba proteus to puncture of its cell membrane is measured. We also exploit the phase stability of SDPM to perform Doppler flow imaging of cytoplasmic streaming in A. proteus. We show reversal of cytoplasmic flow in response to stimuli, and we show that the cytoplasmic flow is laminar (i.e. parabolic) in nature. We are currently investigating the use of SDPM in a variety of different cell types.

In many clinical studies, including those of cancer, it is highly desirable to acquire images of whole tumour sections whilst retaining a microscopic resolution. A usual approach to this is to create a composite image by appropriately overlapping individual images acquired at high magnification under a microscope. A mosaic of these images can be accurately formed by applying image registration, overlap removal and blending techniques. We describe an optimised, automated, fast and reliable method for both image joining and blending. These algorithms can be applied to most types of light microscopy imaging. Examples from histology, from in vivo vascular imaging and from fluorescence applications are shown, both in 2D and 3D. The algorithms are robust to the varying image overlap of a manually moved stage, though examples of composite images acquired both with manually-driven and computer-controlled stages are presented. The overlap-removal algorithm is based on the cross-correlation method; this is used to determine and select the best correlation point between any new image and the previous composite image. A complementary image blending algorithm, based on a gradient method, is used to eliminate sharp intensity changes at the image joins, thus gradually blending one image onto the adjacent 'composite'. The details of the algorithm to overcome both intensity discrepancies and geometric misalignments between the stitched images will be presented and illustrated with several examples.