This PDF file contains the front matter associated with SPIE Proceedings Volume 6849, including the Title Page, Copyright information, Table of Contents, Introduction (if any), and the Conference Committee listing.

We have developed the novel video endoscope imaging techniques; Narrow band imaging (NBI), Auto-Fluorescence Imaging (AFI), Infra-Red Imaging (IRI) and Endo-Cytoscopy System (ECS). The purpose of these imaging techniques is to emphasize the important tissue features associated with early stage of lesions. We have already launched the new medical endoscope system including NBI, AFI and IRI (EVIS LUCERA SPECTRUM, OLYMPUS MEDICAL SYSTEMS Co., Ltd., Fig.1). Moreover ECS, which has enough magnification to observe cell nuclei on a superficial mucosa under methylene blue dye staining, is the endoscopic instrument with ultra-high optical zoom. In this paper we demonstrate the concepts and the medical efficacy of each technology.

A non-invasive, reliable and affordable imaging system with the capability of detecting skin pathologies such as skin cancer would be a valuable tool to use for pre-screening and diagnostic applications. Optical Coherence Microscopy (OCM) is emerging as a building block for in vivo optical diagnosis, where high numerical aperture optics is introduced in the sample arm to achieve high lateral resolution. While high numerical aperture optics enables realizing high lateral resolution at the focus point, dynamic focusing is required to maintain the target lateral resolution throughout the depth of the sample being imaged. In this paper, we demonstrate the ability to dynamically focus in real-time with no moving parts to a depth of up to 2mm in skin-equivalent tissue in order to achieve 3.5&mgr;m lateral resolution throughout an 8 cubic millimeter sample. The built-in dynamic focusing ability is provided by an addressable liquid lens embedded in custom-designed optics which was designed for a broadband laser source of 120 nm bandwidth centered at around 800nm. The imaging probe was designed to be low-cost and portable. Design evaluation and tolerance analysis results show that the probe is robust to manufacturing errors and produces consistent high performance throughout the imaging volume.

In this paper, we present a high performance time-resolved diffuse optical tomography (DOT) system whose working principle and implementation are very different from the conventional methods and features near real-time data acquisition. Dual-wavelength near-infrared light at 784 nm and 808 nm are used as light source. A high bit rate pseudo-random bit sequence (2.488 GB/s) modulates both continuous lights when they go through the external intensity modulator. The diffuse photon density waves responding to such modulated light are detected by array of high sensitivity, high speed silicon avalanche photodetectors (APD). The temporal point spread function can be retrieved very quickly by using hardware auto-/cross-correlation. Our system implementation has achieved system impulse with rise time approaching 0.5 ns which is highly desirable for time-domain DOT application to penetrate depth in few centimeters. The temporal resolution is mainly limited by the photodetectors. System performance is stabilized with accurate temperature control over key components. Experiment results based on tissue-like phantoms verify our design is potential for breast cancer imaging.

Time needed for detection and identification of bacteria can be much shortened using the unique light-scattering pattern after being exposed to the laser source from the new platform named BARDOT (Bacteria Rapid Detection using Optical scattering Technology). The resulting pattern is compared to the compiled pattern library to search for similarity, hence determine the types of bacteria. The system consists of a laser source, an imaging camera, a scattering camera, and a two-dimensional stage. First the imaging camera captures the image of the sample on Petri-dish and locates the center coordinate locations of each cluster. Then the two-dimensional stage translates the Petri-dish such that the incident laser beam is upon the individual sample cluster and performs a centering process which is an fine-adjustment to capture a concentric scattering pattern. The displacement of the platform during this process is determined from the difference of the centroid of the laser beam without sample and that of scattered laser beam with sample. Using MATLAB to design and test the centering algorithm, the time taken for the centering algorithm can be minimized by generating a linear relationship between the lateral distance of the sample movement and the difference of the centroid. The initial algorithm utilized the non-linear relationship without any compensation of the difference of the centroid value. Thus it took multiple steps of motions to reach the center location if the difference of colony center to the laser center is larger than the radius of the sample cluster. With the help of newly designed algorithm, a linear relationship is achieved via identifying the specific location of the starting point of centering algorithm and compensating the corresponding centroid difference to match the actual displacement. Therefore the total time needed to satisfy the centeredness of the scattering pattern is minimized.

Window chambers are support structures implanted in the dorsal skin fold of a rodent model. Optical imaging of window chambers has been used in many basic cancer and vascular biology studies. We have recently shown that this technique can be extended to MRI by using plastic rather than metal window chambers. Here we describe a system for simultaneous optical and MR imaging of the window chambers. It provides many possibilities for independent cross validation of the measurements of one modality from the other. In the system, a GRIN lens images the tissue to the distal end of a coherent imaging fiber bundle, which relays this image to a camera system located outside the magnet room. Both trans- and epi-illumination are provided to this system. Light sources are located outside the magnet room and the light is delivered through fiber optics. A group of fibers are used to deliver white light from under the window chamber for standard transmission imaging, while another single fiber delivers the laser light from the top to induce fluorescence. An appropriate bandpass emission filter is inserted between the lenses at the camera end for fluorescence imaging. Results of simultaneously optical and MR imaging of tumor and vessel are presented.

A multimodal imaging system has been developed for tooth tissue imaging. This imaging system is designed to obtain one or more two-dimensional images of the tooth tissue, and those two-dimensional images are rendered with advanced algorithms to provide a high-contrast image. This system combines polarized reflectance imaging, fluorescence imaging, and optical coherence tomography (OCT) imaging. The imaging system design, as well as some experimental results, will be discussed in the presentation.

We present human eye wavefront generator (HEWG) introduced inside aberrometer for dynamic reproduction of human eye aberrations. It's main element is bimorph deformable mirror and a telescope. Deformable mirror generates human eye aberrations in real time. We have recorded aberrations time traces for plurality of subjects using aberrometer and reproduced them with the help of the generator at 10Hz frequency, that is inherit to human eye aberrations dynamics. Experimental results indicate that HEWG can reproduce dynamics of human eye aberrations with residual error less than λ/10 microns. Such a model can be useful for testing, for example, customized contact lenses or wavefront guided aberrometers.

There have been several technologies to enable high resolution cross-sectional images of biological tissues in optical coherence tomography (OCT) method. Optical frequency comb (OFC) source has been proposed to overcome the crosstalk problem among the CCD detector pixels of the continuous spectrum of light source. Recently, a passive-type OFC is demonstrated simply placing a Fabry-Perot interferometer filter right after the broadband light source, but it shows a high loss of output light power and limited tenability of channel spacing of multi-wavelength. In this work, we experimentally demonstrate a spectral comparison of a novel multi-wavelength source based on a fiber Sagnac interferometer. The channel spacing is flexibly tuned by the effective length control of polarization-maintaining fiber (PMF). The uniform and stable multi-wavelength spectral distribution is also helpful to obtain the higher sensitivity from the lower exposure intensity source to get a better quality spectral OCT image.

We evaluate the performance of ODISsey^TM Tissue Oximeter (ViOptix, Inc., Fremont, CA) against co-oximeter. Concurrent oxygen saturation measurements were made in three dog limbs surgically removed and perfused with an extracorporeal blood circulation system. Oxygen saturation was adjusted in steps ranging from 95% down to 5% as monitored by the co-oximeter. The co-oximeter was used to measure the oxygen saturation of the whole blood drawn from both the arterial and the venous ports of the limb. The tissue oxygenation measured by the ODISsey^TM tissue oximeter was compared with the average of the arterial and the venous blood oxygenation measured by the co-oximeter. Linear correlation was observed between the average oxygenation given by the co-oximeter and the ODISseyTM readings, with a root-mean-square difference of 7.6% and the correlation coefficient of 0.941, calculated from N = 194 data points.

The association of a beveled collection fiber with a half ball lens is here simulated in order to illustrate the enhancement brought by this simple method in depth resolved measurements, the latter being of major importance in the study of epithelial tissues. Indeed, in order to follow the carcinogenesis stage by a non invasive technique such as fluorescence or Raman spectroscopy, it is valuable to be able to discriminate the depth origin of the detected signal. The proposed probe design consists in a beveled collection fiber positioned on the center of the half ball lens flat surface with an off-axis excitation flat fiber. Depending on the bevel angle of the collection fiber, discrimination between the epithelial and the stromal layer can be realized.

Tissue optical properties at ultraviolet A (UVA) and visible (VIS) wavelengths are needed to elucidate light-tissue interaction effects and optimize design parameters for spectroscopy-based neoplasia detection devices. Toward the goal of accurate and useful in vivo measurements, we have constructed and evaluated a system for optical property measurement at UVA-VIS wavelengths. Our approach involves a neural network-based inverse model calibrated with reflectance datasets simulated using a condensed Monte Carlo approach with absorption coefficients as high as 80 cm^-1 and reduced scattering coefficients as high as 70 cm^-1. Optical properties can be predicted with the inverse model based on spatially resolved reflectance measured with a fiberoptic probe. Theoretical evaluation of the inverse model was performed using simulated reflectance distributions at random optical properties. Experimental evaluation involved the use of tissue phantoms constructed from bovine hemoglobin and polystyrene microspheres. An average accuracy of ±1.0 cm^-1 for absorption coefficients and ±2.7 cm^-1 for reduced scattering coefficients was found from realistic phantoms at five UVA-VIS wavelengths. While accounting for the very high attenuation levels near the 415 nm Soret absorption band required some modifications, our findings provide evidence that the current approach produces useful data over a wide range of optical properties, and should be particularly useful for in vivo characterization of highly attenuating biological tissues.

Photothermal therapy employing nanomaterials is a promising approach to selectively treat targeted tissues with abnormal characteristics such as tumors. While vital research has focused on the use of these materials in biomedical applications, net effects of these materials in biological environments are still not well understood. For reliable biomedical applications, it is crucial to quantitatively evaluate thermal properties of these materials in biological and physiological environments. To this end, we have developed a highly integrated measurement platform and examined local thermal properties of single gold shell nanocrystals in biomimetic environments. These nanoshells consist of a silica core with an outer gold coating. For quantitative measurement of the local thermal profile of gold nanoshells, we monitor lipid phase transitions triggered by gold nanoshell thermal excitation. Dried lipid layers with adsorbed gold nanoshells were placed in an aqueous environment. Photothermal excitation of the gold nanoshells induced localized liposome budding as the lipids were raised above their transition temperature. Single particle tracking of gold nanoshells in solution and within liposomes revealed larger diffusion rates for the confined nanoparticles, likely due to a raised local temperature.

This study examines the accuracy of the Living Image® Software 3D Analysis Package (Xenogen, Alameda, CA) in reconstruction of light source depth and intensity. Constant intensity light sources were placed in an optically homogeneous medium (chicken breast). Spectrally filtered images were taken at 560, 580, 600, 620, 640, and 660 nanometers. The Living Image® Software 3D Analysis Package was employed to reconstruct source depth and intensity using these spectrally filtered images. For sources shallower than the mean free path of light there was proportionally higher inaccuracy in reconstruction. For sources deeper than the mean free path, the average error in depth and intensity reconstruction was less than 4% and 12%, respectively. The ability to distinguish multiple sources decreased with increasing source depth and typically required a spatial separation of twice the depth. The constant intensity light sources were also implanted in mice to examine the effect of optical inhomogeneity. The reconstruction accuracy suffered in inhomogeneous tissue with accuracy influenced by the choice of optical properties used in reconstruction.

To improve efficiency and diagnostics, an optical testbench was developed for routine assessment of the optical quality of endoscopes during their lifespan. An endoscope is positioned on an optical bench looking at a small LCD screen showing computer generated test targets. Images through the endoscope are captured by a high-resolution USB-2 camera, analyzed with dedicated software and stored in a database. Additionally, the transmission of the illumination fibers of the endoscope is determined using a calibrated LED light source and a photo cell. The database enables the tracking of the degradation of each endoscope over time and comparison with other endoscopes of the same type. Using the data collected and the optical quality of new endoscopes as reference, criteria are developed to reject endoscopes at some point in their lifespan. Experience with the optical testbench over the last two years shows that: 1) Optical quality can be measured reproducible over time, 2) degradation in optical quality is often gradually, but may also occur suddenly, and 3) there is a large variation (>20%) in optical quality of new or repaired endoscopes. Routine quantitative assessment of the optical quality of endoscopes will contribute to the efficiency and the quality of surgery.

We studied the performance of an OCT imaging modality on the task of detecting an abnormality in biological tissue. Optical propagation in biological samples is dominated by scattering due to fluctuations in refractive index. We used the first order multiple scattering approximation to describe the scattered field from the tissue. The biological tissue was described by it permittivity field and the corresponding scattering potential. The normal state of the tissue (the background) was modeled as a spatial Poisson field of randomly distributed scattering centers, and the abnormality (the target) as a region with a higher concentration of scattering centers embedded in the background. The target detectability was then calculated using a quadratic observer. We considered the effect of fluctuations from the broadband source, the shot noise fluctuation of the imaging system, and the scattering noise due to refractive index fluctuation in the biological tissue. We also studied the detectability of an embedded abnormality in biological tissue with respect to to size of th abnormality.

Blood movement inside a finger causes color changes in a fingerprint image during a fingerprint input action. We consider models to relate this information to the stiffness of a collective blood vessel in a finger. In the simple resistor model, finger area is regarded as voltage and the color change was regarded as current. We define its resistance as the ratio of these two quantities according to Ohm's law. However, experiments show that this resistance value increases as the force applied to a finger increases. Second, we consider the variable resistor model to account for this force dependency. We assume that the driving force decreases the cross-section of the blood vessel and that stress and strain are related by Young's modulus. Experiments by six participants show reasonably good fittings for the extracted signals and the model's predictions. The correlation coefficient between the extracted parameter and the average blood pressures of the participants was 0.75.