Fluorescence Imaging is an experimental clinical technique for tumor detection, which has been gaining interest over the last years.

An optical imaging and spectroscopy system has been developed for the study of in vivo fluorescence of nasopharyngeal tissue through an endoscope. The system records the fluorescence signal in the imaging plane of the endoscopic system. This allows analyze the characteristics of the light induced fluorescence (LIF) spectra recorded by each pixel of the 2D detector which may be used for fluorescence endoscopic imaging. If the endoscope for fluorescence endoscopy is the same as one employed for the in vivo fluorescence study, the algorithms developed to distinguish the diseased tissue from normal tissue based on the in vivo fluorescence study should be highly reliable for fluorescence imaging of lesions. In this work, fluorescence spectra were collected from 27 full term patients. Different algorithms were tested for separation of cancerous lesions from normal tissue. High sensitivity and specificity were achieved.

The steady state native fluorescence emission and excitation spectra of human normal and cancerous oral tissues are studied in the visible region. The fluorescence excitation spectrum is recorded for 600 nm emission by scanning the excitation. The excitation spectrum of normal tissues has peaks at 406, 524 and 552 nm, whereas the cancerous tissues have peaks at 406, 513 and 552 nm respectively. The fluorescence emission spectra were also recorded at 405 and 560 nm excitations. The emission spectrum of cancerous tissues has two distinct peaks at 604 and 660 nm. It is also observed that there is a distinct difference between normal and cancerous tissues at 560 nm excitation. The ratio parameter R_{1$ equals (I(subscript 406}/I₅₅₀) is introduced from the excitation spectrum for 600 nm emission and two ratio parameters R₂ equals (I₄₇₀/I₆₀₀) and R₃ equals (I₄₇₀/I₆₆₀) are introduced for the emission spectrum at 405 nm excitation. Among the three ratio parameters the R₁ classifies the normal and cancerous tissues at a specificity and sensitivity of 83 percent and 93 percent respectively. A critical value of 1.8 is suggested for classifying the normal from cancerous tissues.

Rayleigh light scattering has not yet been used for quantitative investigations of heterogeneous systems. Preconditions such an experiment are a well defined scattering geometry and independent information about the local state of the sample. We have designed a new instrument that meets these criteria: a light-scattering microscope with simultaneous imaging. We demonstrate the ability to characterize local differences within one tissue type as well as global differences between tissue types. Real space images of the sample are taken by normal video microscopy techniques. The light scattering pattern in analyzed by the evaluation of wave-vector dependence and scattering direction of the scattered intensity. Statistical analysis of scattering patterns show what is important for the characterization and classification of tissues and heterogeneous structures. Real space images provide context for scattering analysis. The light scattering microscope is a powerful tool for characterization of local structural order in inhomogeneous structures like tissues.

As part of our ongoing efforts to understand the fundamental nature of light scattering form cells and tissues, we present data on elastic light scattering from isolated mammalian tumor cells and nuclei. The contribution of scattering from internal structures and in particular from the nuclei was compared to scattering from whole cells. Roughly 55 percent of the elastic light scattering at high- angles comes from intracellular structures. An upper limit of 40 percent on the fractional contribution of scattering form cells in tissue was determined. Using cell suspensions isolated from monolayer cultures at different stages of growth, we have also found that scattering at angles greater than about 110 percent was correlated with the DNA content of the cells. Based on model calculations and the relative size difference of nuclei from cells in different stages of growth, we argue that this difference in scattering results form changes in the internal structures of the nucleus. This interpretation is consistent with our estimate of 0.2 micrometers as the mean size of the scattering centers in cells. Additionally, we find that while scattering from the nucleus accounts for a majority of internal scattering, a significant portion must result from scattering off of cytoplasmic structures such as mitochondria.

We have developed a Monte Carlo algorithm that calculates all sixteen, 2D elements of the diffusing back scattering Mueller Matrix for highly scattering media. Using the Stokes-Mueller formalism and scattering amplitudes calculated with Mie theory, we are able to consider polarization dependent photon propagation in highly scattering media. The numerically computed matrix elements are compared to experimental data obtained from particle suspensions with different particle sizes and fibroblast cell suspensions. The numerical results show good agreement in both azimuthal and radial direction with the experimental data, and suggest that in the fibroblast suspensions subcellular structures with a typical size of 200 to 300 nm dominate the back scattering behavior.

We have used the cross-polarization method to eliminate the specular reflection and enhance the diffusive back- scattering of polarized fluorescence excitation light from the turbid media. The image of the ratio between fluorescence and cross-polarized reflection provides a map of the fluorescence yield, define das the ratio of the autofluorescence emission to the illumination incidence, over the surface of the turbid media when excitation and collection of fluorescence are highly inhomogeneous. The simple ratio imaging technique shows the feasibility to detect early cancer, which usually starts from the superficial layer of tissue, based on the contrast in the fluorescence yield between lesion and normal tissue.

The aim of this effort is to test the feasibility to develop endoscopic optical imaging technology for cancer screening inside the human body. The degree of success of such optical imaging instrumentation depends on the presence of optical differences among the various tissue components of interest of a body part or organ. Our research approach involves the utilization of the spectral polarization difference imaging technique incorporated into endoscopic imaging modalities. To establish methodology for optimum operation, we have built a compact imaging system for in vitro studies of human tissue samples of interest in a clinical environment.

Elastic scattering or diffuse reflectance spectroscopy offers the possibility of distinguishing between normal and neoplastic tissue with a relatively simple optical measurement. The measurement of the reflection of light has previously been shown to be sensitive to the size and distribution of both intra and inter-cellular structures as well as absorption from chromatophores which are present in the tissue. By coupling a white light source and spectrometer to optic fibers it is possible to construct probes which can be inserted precutaneously or intra- operatively into breast tissue or which can pass down the channel of an endoscope and take in-vivo spectra of diseased and normal tissue in the Gastro-Intestinal tract. Spectra are reported from a large number of patients with a variety of benign, metaplastic, dysplastic and cancerous conditions. Some differences that have been observed in these spectra are discussed and the merits and disadvantages of 'optical biopsy' as an in-vivo diagnostic tool are examined. It is shown that to a relatively high degree of sensitivity and specificity it is possible to distinguish cancerous from normal tissue in a number of cases. The methods of distinguishing spectra and some possible modalities for their improvement are discussed.

Several dyes are currently used for various biomedical applications due to their biocompatibility and high molar absorptivity. Localization of dyes in tumors may be mediated by several factors such as leaky vasculature and high metabolic activity in proliferating cells. However, these mechanisms of action make it difficult to differentiate inflammation from benign or malignant tumors. In order to enhance their tumor specificity, dyes have been conjugated to biomolecules that target unique factors in various diseased state. However, such large biomolecules can elicit adverse immunogenic reactions in humans, and are often preferentially taken up by the liver. Furthermore, for solid tumors which may rely on diffusion of the biomarkers from the vascular, penetration of large dye conjugates is not favorable. To overcome these problems, we designed and synthesized novel dye-peptide conjugates that are receptor specific. The efficacy of these new fluorescent contrast agents was tested in vivo in well-characterized rat tumor lines. The resulting optical images demonstrate that successful specific tumor targeting was achieved.

We demonstrate the nanosecond time-gated spectroscopy of plume in laser ablation of biological tissue, which allows us to detect calcium (Ca) with high sensitivity by the use of either a UV or a near-IR laser pulse. Clear and sharp peaks of Ca⁺ appear in the luminescence spectrum of laser-ablation plume although the Ca content is only 0.1 percent in human hair and nail. Luminescence peaks of sodium atom (Na) and ionized carbon are also detectable. This specific spectroscopy is low invasive because a single low-energy laser pulse illuminates the tissue sample, and it does not require any poisonous sensititizers like fluorescence dye. This method, therefore, is a promising candidate for optical biopsy in the near future. In particular, Ca detection of human hair may lead to new diagnosis, including monitor of daily intake of Ca and a screening diagnosis of osteoporosis.

The relative proportions of genetically distinct collagen types in connective tissues vary with tissue type and change during disease progression, development, wound healing, aging. This study aims to 1) characterize the spectro- temporal fluorescence emission of fiber different types of collagen and 2) assess the ability of time-resolved laser- induced fluorescence spectroscopy to distinguish between collagen types. Fluorescence emission of commercially available purified samples was induced with nitrogen laser excitation pulses and detected with a MCP-PMT connected to a digital storage oscilloscope. The recorded time-resolved emission spectra displayed distinct fluorescence emission characteristics for each collagen type. The time domain information complemented the spectral domain intensity data for improved discrimination between different collagen types. Our results reveal that analysis of the fluorescence emission can be used to characterize different species of collagen. Also, the results suggest that time-resolved spectroscopy can be used for monitoring of connective tissue matrix composition changes due to various pathological and non-pathological conditions.

We have used resonance Raman scattering as a novel, non- invasive, in-vivo optical technique to measure the concentration of carotenoid pigment in the human retina. Using argon laser excitation we are able to measure two strong carotenoid resonance Raman signals at 1159 and 1525 wave numbers, respectively. The required laser power levels are within the limits given by safety standards for ocular exposure. Of the approximately ten carotenoid pigment found in normal human serum, the species lutein and zeaxanthin are concentrated in high amounts in the cells of the human macula, which is an approximately 5 mm diameter area of the retina in which the visual acuity is highest. These carotenoids give the macula a characteristic yellow coloration, and it is speculated that these molecules function as filter to attenuate photochemical damage and/or image degradation under bright UV/blue light exposures. In addition, they are thought to act as free-radical scavenging antioxidants. Studies have shown that there may be a link between macular degenerative diseases, the leading cause of blindness in the elderly in the US, and the presence or absence of the carotenoids. We describe an instrument capable of measuring the macular carotenoids in human subjects in a non-invasive, rapid and quantitative way.

Global analysis of time-resolved fluorescence measured at multiple emission wavelengths was applied to simulated fluorescence spectra and arterial fluorescence spectra. Fluorescence of human optic samples was produced with nitrogen laser excitation. Simulated spectra had decay characteristics in the range expected from previous studies of artery tissue. For both types of spectra, the emission decay was analyzed with global analysis to model the decay with a sum of exponentials. Decay constants were held fixed across wavelengths while pre-exponential coefficients were wavelength-dependent. For the simulated spectra, global analysis was compared to the traditional method in which decay constants and pre-exponential coefficients are assumed wavelength-dependent. On the simulated data, three decays could be reliably estimated by global analysis seven when only two exponential decays were identified with the traditional method. On the arterial data, the intermediate decay and the long decay significantly increased between normal samples and fibrous plaque. The pre-exponential coefficient of the long decay was larger in the blue rang of the spectrum for the samples with advanced atherosclerosis. We conclude that global analysis markedly improves the recovery of exponential decay trends in time-resolved fluorescence spectra. Application to artery tissue fluorescence reveals characteristic spectral changes associated with atherosclerosis.

Most current configurations for optical biopsy contain fiber optic bundles at both the delivery and receiving ends of the optical system. Some layouts include distal lenses to either collimate or focus the incident light at various depth locations across the tissue. The inherent beam divergence, along with the highly scattering nature of the living tissue, are known to limit the penetration depth of the probe and the spatial or temporal resolution of the detected signal. In this work we study a novel modality for tissue illumination based on the use of long-range nondiffracting beams (LRNB). LRNB represent narrow-width light pencils with a constant or linearly varying axial intensity that propagate over large distances without diffractive spreading. Recent tests have demonstrated that LRNB exhibit in significant intensity distortions when operated as beacon beams through atmospheric turbulence. Our numerical and software simulations show that LRNB may offer the potential for larger penetration depth and enhanced contrast over setups using conventional laser beams. Clinical applications include diagnosis, laser surgery and photodynamic therapy.

We considered the limited number of light-induced fluorescence applications for marketed ultra-bright blue LEDs where they can compete with versatile laser sources. Satisfactory optical output and miniature size as well as low power consumption of blue LEDs emitting at 470 nm allow to consider them as a promising alternatives to metal vapor or gas lasers used in many expires LIF applications. Available to authors LEDs form Hewlett-Packard, Micro Electronics Corp., Nichia Chemical Industries Ltd. and Toyoda Gosei Co. were tested to comply with demands to a tissue excitation source for portable spectroscopes. The optical performance of LEDs has shown that selected group of InGaN LEDs could be successfully used for that. The miniature illuminator that includes LED, focusing condenser, filter set and distal fiberoptic light concentrator was designed and tested in conjunction with portable CCD- equipped spectroscope. Operating in dark condition the proposed LED illuminator provides the level of fluorescence signal sufficient to detect spectral abnormalities in human Caucasian skin and excised gastrointestinal samples. All tissue autofluorescence data taken under LED illumination were compared with readings under He-Cd laser excitation and showed a good match. A new diagnostic designs based on LEDs were considered for clinical use.

Based on the microscopic properties of colonic tissues, a five-layer colon optical model was developed to calculate the excitation light distribution in the tissue and the fluorescence escape function from the tissue by Monte Carlo simulations. The theoretically modeled fluorescence spectrum fits well to the experimental results, demonstrating that the microscopic properties of tissue applied in the colon optical model can be quantitatively correlated with the macroscopic autofluorescence measurements.

Fluorescence images of ex vivo head and neck tissues were acquired at multiple combinations of emission and excitation wavelengths. The wavelength combinations were selected to map different tissue molecules and structures whose fluorescence signatures have bene used to detect cancer. Fluorescence maps were generated by ratioing fluorescence image intensities. These ratio maps enhanced the ability to recognize regions of tumor and other features in tissues. Histopathological analysis was performed on the tissue samples. Location and shape of features observed in the fluorescence images were correlated with structures observed in histopathology.

The experimental study of mechanical stress distribution in biological tissues in-vivo is of interest for some biomedical applications. This work considers the problem of light and stressed tissue interaction and the inverse problem of stress field reconstruction in 2D and 3D cases. Optical tomography is one of the most promising methods of solving these problems. This technique involves the reconstruction of the refractive index field using the measurement of waveform distortion. The reconstruction of stress field requires establishing the relation of the stress tensor to the variation of refraction index. A simple photoelastic model is a reasonable first approximation due to normal functioning of biological tissue. The simple photoelastic model is a reasonable first approximation due to normal functioning of biological tissue. The propagation equation that describes the light propagation through the optically active elastic media obtained in the solving of the forward problem in terms of geometric optics approach. Interferometric, shlieren and depolarization methods of experimental data acquisition are considered. In general, 3D state of a stressed tissue should be described by six components of the stress tensor, but only three propagation equations appear to be independent. To close the system of equations, we have used three partial differential equilibrium equations with appropriate boundary conditions. The system of equations of interferometric tomography is studied in detail. In this case, the separation of stress tensor components results from analytical solving in the Radon domain. For special case of 2D deformation we need only one propagation equation and two equilibrium equations. It is shown that 3D problem can not be reduced to 2D problem in the general case of tensor field tomography. This sends us in search of special 3D algorithms. The use of wavelets is one of perspective ways of tomographic reconstruction under strong noise. 2D and 3D algorithms of the inverse Radon transform through inverse wavelet transform in noisy conditions have been developed.

We describe a non-invasive method for the determination of optical parameters of highly scattering media, such as biological tissue. An advantage of this method is that it does not rely on diffusion theory, thus it is applicable to strongly absorbing media and at small source-detector separations. Monte Carlo simulations and phantom measurements are used to illustrate the achievable accuracy of the system. The method was applied to non-invasive in- vivo tracking of haemoglobin concentration in biological tissue. The results correlated well to clinically determined Hb concentrations.

Light propagation in a turbid medium such as the skin tissue depends on both the bulk optical properties and the profiles of the interfaces where mismatch in the refractive index occurs. In this paper we present recent results of investigations on the light distribution inside a human skin tissue phantom for an incident converging laser beam and its dependence on the roughness of the interfaces. The human skin tissue is modeled by a two-layer structure with a thin layer of epidermis on top of the dermal layer. Within each layer, the tissue is considered macroscopically homogeneous and the two interfaces, between ambient medium, epidermis and dermis, are treated as random rough surfaces. The distribution of laser light with wavelength near 1 micrometers in the tissue phantom is considered using a recently developed method of Monte Carlo simulation. The dependence of the light distribution on the surface roughness and index mismatch are presented, and their relevance to the possible laser surgery under skin surface and the measurements of optical properties of the skin tissues is discussed.

The weak absorption of near-IR light by skin tissue has offered an important optical window for diagnostic possibilities with optical means including optical biopsy. The strong scattering of the near-IR light by skin tissue, however, presents a great challenge to the modeling of light propagation through and the optical measurement of the tissue. We have measured transmittance and reflectance of fresh porcine skin and performed Monte Carlo simulations to inversely determine the absorption coefficient, scattering coefficient and asymmetric coefficient of tissue samples in the spectral range from 900 to 1500nm. The state of cellular integrity following optical measurements was verified using transmission electron microscopy. These results were correlated with the possible effects on the measurements of the tissue optical properties.

Spatial variation of fluorescence intensity in human breast tissues was studied. A diffusion theory model describing fluorescence light energy distribution inside a thick and turbid medium like human tissue was developed. Experimental data were fitted with the theoretical model to obtain reduced scattering coefficient and absorption coefficient at different fluorescence wavelengths with 488nm excitation of normal, malignant and benign breast tissues. Computed reduced scattering and absorption coefficients of malignant tissues were observed to be more as compared to the normal tissues and benign tumors.

Fluorescence images were acquired form gynecological tissues for multiple combinations of emission and excitation wavelengths in the UV and blue spectral regions. The wavelength combinations were selected to highlight different tissue molecules whose fluorescence signatures have been shown to potential in the detection of malignancy. These images were analyzed to determine the size, shape and location of different tissue structures.

An approach to computer assisted imaging diagnosis is introduced for tumor detection in tissues, using softer light sources and detection from both transmission and reflection. A 3D model, described by the diffusion equation, is developed to simulate the photon propagation inside human tissue to detect the tumor or other inclusions. The inclusion is modeled with a different absorption coefficient form that of the otherwise homogeneous media. The ADI algorithm is used to solve the diffusion equation on each spatial grid at each time instance in an interactive manner. The data on a cut section of the model at a certain moment is visualized to be a picture according to the photon intensity value. Pictures of different moments can be linked to construct a dynamic movie for the presentation of the photon propagation procedure on a specific cut-plane. The movies show that in the homogeneous model the photon density has a symmetric concentric distribution on the plane opposite the source. When an off-center inclusion is introduced inside the model, a clear asymmetry appears. The shadow of the inclusion is determined by both the size and the location of the inclusion. If the inclusion is moved to the center, the symmetry is restored, though different from the original image. To reach a more precise result, the derivative of the intensity, with respect to the distance from the light source is computed. From the dynamic derivative curves, the size of the inclusion can be estimated. Experiments verify the conclusion on inclusions of arbitrary dimensions which are larger than 5 mm. A scanning search strategy is proposed for unknown inclusions.

Near IR laser is being explored in clinic diagnosis of early-stage cancer. Diffusion equation is used as the mathematical model to describe photon propagation inside human tissues. In this paper, two numerical algorithms, ADI and FEM, are applied in solving the diffusion equation. All the algorithms reach a satisfactory precision on a human- tissue model of realistic size. Results form simulation and experimental are compared and show a good match. Further analysis on algorithm convergence for both spatial grid size and time step also shows the algorithmic stability. The multigrid version of both ADI and FEM are developed to save computational time and memory. The multigrid algorithms use fine grid size in the region of interest and coarse grid size elsewhere. Parallelization implementation is completed for all the algorithms in both share-memory mode and message passing mode. Numerical simulation experiments show that all simulators can serve as computed experiments, i.e. an alternative to physical experiments.

A novel approach to the quantitative image reconstruction in diffuse optical tomography is proposed. The special structure of the transport equation is used to formulate the iterative image reconstruction algorithm as a process updating the estimates of the optical properties from the solution of an intermediate tomographic problem The ability of the technique to reconstruct simultaneously maps of both absorption and reduced scattering coefficients in 2D geometry is demonstrated using simulated frequency-domain data. The potential advantages of the new approach include its ability to fully retain the non-linear character of the inverse problem while at the same time avoiding either gradient or Jacobian calculations and eliminating the need in an additional regularization mechanism.

Sized-fiber array spectroscopy describes a device for measuring the absorption and reduced scattering properties of tissue. The device consists of two fibers with different diameters that measure the amount of light back-scattered into each fiber. Only one fiber emits and collects light at a time. Recent innovations allow for spatially limited measurement diffuse reflectance over a wavelength range of 500-800 nm. Reflection spectra of in vitro an din vivo porcine tissue are presented for a device with 200 and 600 micrometers fibers to demonstrate its performance.

The validity of self-reported skin color was assessed by comparing the responses of a skin color survey with the external measure of diffuse reflectance spectrophotometry. Reflectance spectra of 108 subjects were measured at sites on the arm normally exposed to sunlight and sites normally unexposed to sunlight. The reflectance spectra were analyzed with a variety of discriminating algorithms, such as principal component analysis, and competitive neural networks. For subjects with light and dark skin, there was good correspondence between the survey results and groupings derived by the neural network analysis. For those people reporting medium skin color, the correspondence with the neural network groupings was poor. It was unclear if this was due to poor self-reporting or deficiencies in the spectral analysis.

Fluorescence properties of flavins in normal, benign and malignant human breast tissues have been investigated between 500-700 nm using 488 nm excitation of an Ar-ion laser. The combination of fluorescence anisotropy and spectral profiles can distinguish normal, benign and malignant from one another. The fluorescence spectra may be characterized by two major bands with the width of the band at 580nm being the distinguishing parameter. The polarization study of human breast tissues shows higher anisotropy for the tumor tissues compared to their normal counterparts. The effects of multiple scattering on depolarization of fluorescence is confirmed by polarization measurements on thin tissue sections which show higher anisotropy values compared to the thick samples. An important observation is the increase in the difference of anisotropy values between some of the normal and malignant tumor samples while going form thick to thin tissues.

Biopsy or photo dynamic therapy of tumors are usually investigated by fluorescent diagnostics methods. Information on modified method of fluorescence diagnostics of inflammatory diseases is represented in this research. Anaerobic micro organisms are often the cause of these pathological processes. These micro organisms also accompany disbiotic processes in intestines.

Experimental study and computer simulation of light propagation in turbid media with different concentration of scattering particles are described. Gel phantoms with adding of Intralipid-20 percent as a model of living tissue have been investigated. Experimental results have shown existing of transmittance and reflectance saturation in dependence on volume fraction of the scatterers. The inverse adding- doubling method and Rayleigh-Gans approximation of the Mie theory were applied to describe the light propagation. The explanation of the experimental results is presented.

Human prostate in-vitro tissues were studied using near IR spectral polarization imaging. Different imaging methods using the light scattered or emitted from prostates tissues and contrast agents were performed on various model samples, which consisted of a small piece of absorber or prostate tissue dyed with indocayanine green embedded inside a large piece of prostate tissue at different depths. Small foreign objects with a diameter of approximately 1 mm hidden inside the host prostate tissues at depths of 3.0 mm, 4.5 mm, and 8.5 mm were imaged and identified using the scattering light, tissue emission wing and contrast agent emission light imaging methods, respectively.

Normal, fibroadenoma, malignant, and adipose breast tissues were investigated using Kubelka-Munk Spectral Function (KMSF). The spectral features in KMSF were identified and compared with absorption spectra determined by transmission measurements. A specified spectral feature measured in adipose tissue was assigned to (beta) -carotene, which can be used to separate fat form other molecular components in breast tissues. The peaks of (KMF) at 260nm and 280nm were attributed to DNA and proteins. The signal amplitude over 255nm to 265nm and 275nm to 285nm were found to be different for malignant fibroadenoma, and normal tissues.