Currently breast cancer diagnosis is made clinically through triple assessment: annual clinical breast examination, x-ray mammography or breast ultrasound imaging, and biopsy. The majority of women with suspicious breast lesions undergo either stereotactic (needle) or surgical (excisional) biopsy. Due to a high incidence of "false positives" at clinical breast diagnosis and "false negatives" at surgery, a large number of women undergo unnecessary and costly breast surgery. We describe our program of development of techniques and instrumentation for clinical application of NIR Raman spectroscopy for improved breast cancer diagnosis.

Abstract: laboratory Raman spectroscopy was performed on 59 lymph node sections from breast cancer patients, demonstrating 91% sensitivity and 93% specificity for the correct classification of positive node spectra in a model.

In recent years, the use of Raman spectroscopy for the detection and diagnosis of disease has steadily grown within the research field. However, this research has primarily been restricted to oncology. This research expands the use of Raman spectroscopy as a potential tool for the diagnosis of Alzheimer's disease, which is currently the most prevalent, and fastest growing type of dementia in the Western world. Using a commercial Raman spectrometer (Renishaw PLC ®, UK) flash frozen post-mortem ex vivo brain tissue sections were illuminated using a high power (20mW) 830 nm near infrared diode laser, and subsequently spectra were gained in the region of 2000–200 cm^-1 from a 10 second accumulation time. Ethical approval was gained for the examination of 18 individual donors exhibiting varying states of Alzheimer's disease, Huntingdon's disease and their corresponding age-matched healthy controls. Following on from previous preliminary studies, the Raman spectra were found to be highly reproducible, and when examined further, the spectra showed differences relating to the content and structure of the proteins in the individual brain samples, in particular, the beta-amyloid protein structure found in Alzheimer's disease patients. Principle components analysis further determined these protein structural changes, with Alzheimer's disease and Huntingdon's disease samples being defined from the healthy controls, and from each other.

This study evaluated the diagnostic ability of near-infrared (NIR) Raman spectroscopy for identifying the malignant tumors from normal and benign tissues in the colon. In this work, a rapid NIR Raman system was utilized for tissue Raman studies. A total of 105 colonic specimens were used for Raman studies, including 41 normals, 18 polyps (benign), and 46 malignant tumors (22 of moderately differentiated adenocarcinomas and 24 of poorly differentiated adenocarcinomas). The results showed that high-quality Raman spectra in the 800-1800 cm^-1 range can be acquired from human colonic tissues in vitro, and Raman spectra differed significantly between normal and malignant tumor tissue. The diagnostic algorithm using the Raman intensity ratios of I_1085/1445 vs. I_1002/1445 can yield a diagnostic sensitivity of 100% and specificity of 96.6% for differentiation between between normal, benign and malignant colonic tissue. This work demonstrates that NIR Raman spectroscopy technique has a significant potential for the noninvasive diagnosis of colon cancer in vivo based on the evaluation of changes of molecular vibrations of biomolecules in tissue.

A recent developed pattern recognition algorithm, Support Vector Machines (SVM), was employed to classify nearinfrared Raman spectroscopy data collected from normal and cancerous ENT tissues. Three types of classifiers, linear, 3^rd order polynomial, and radial basis function, were used. Highest diagnostic accuracy was obtained by 3^rd order polynomial with a sensitivity of 91.86% and a specificity of 100%. The possibility to simplify SVM implementation was also explored by using principal component analysis (PCA) to extract significant principal components. It was found that the first five principal components as the data inputs were already sufficient to produce sensitivities of 100% and specificities of 100% for all these three classifiers. Combination PCA and linear discriminant analysis (LDA) to classify these ENT data was also performed and analysis results show that both methods, combination PCA & SVM and PCA & LDA yielded comparable performance.

Autofluorescence (AF) bronchoscopy is a useful tool for early cancer detection. However the mechanisms involved in this diagnosis procedure are poorly understood. We present an in vivo autofluorescence imaging study to access the depth of the principal contrast mechanisms within the bronchial tissue comparing a narrow band and broad band violet fluorescence excitation. Knowledge of this parameter is crucial for the optimization of the spectral and optical design of clinical diagnostic AF imaging devices. We observed no differences in the chromatic contrast using the two excitation modes, indicating that the principal contrast mechanisms have a non-superficial character.

To detect bronchial carcinoma by autofluorescence, we measured in-vivo, in an in-vivo model, and in-vitro the spectra of tumor and normal tissue by a fiber-optic-spectrometer. The main difference between tumor and bronchial tissue is the intensity of the 505 nm main peak.

The aims of this work are investigation and differentiation between normal skin and pigmented skin lesions by using optical reflectance spectroscopy. Optical reflectance spectra in the wavelength range 400-900 nm were obtained from malignant and benign skin lesions and characteristic differences between them were studied. The lesions were preliminarily classified dermatoscopically (MoleMax II, DERMA Instruments). All suspicious lesions were excised and the materials were investigated histologically by standard methods. An algorithm was developed for differentiation and valuation of the skin tissue condition using dimensionless ratios of the reflectance signal intensities of normal skin and pigmented lesions, which allowed the use of non-normalized reflectance spectra for lesion assessment. These ratios were found to have a definite diagnostic potential, as significant differences were observed between normal skin, benign and malignant pigmented lesions.

We present the results of a clinical study using ESS to detect dysplasia in the esophagus. We focus on the use of novel statistical techniques and the clinical benefits this technique provides.

Sentinel node biopsy is the new standard for lymphatic staging of breast carcinoma. Intraoperative detection of sentinel node metastases avoids a second operation for those patients with metastatic lymph nodes. Elastic scattering spectroscopy is an optical technique which is sensitive to cellular and subcellular changes occurring in malignancy. We analyzed 2078 ESS spectra from 324 axillary sentinel nodes from patients with breast carcinoma. ESS was able to detect metastatic lymph nodes with an overall sensitivity of 60% and specificity of 94%, which is comparable to existing pathological techniques. Nodes completely replaced with metastatic tumour were detected with 100% sensitivity, suggesting that further improvement in sensitivity is likely with more intensive optical sampling of the nodes.

We have applied VIS-NIR diffuse reflectance spectroscopy to study different human skin lesions. A new set of features has been derived through the analysis of their spectra to discriminate among normal skin and skin lesions.

Development of new optical sensor technologies has a major impact on the progression of diagnostic methods. Specifically, the optical analysis of breath is an extraordinarily promising technique. Spectroscopic sensors for the non-invasive ¹³C-breath tests (the Urea Breath Test for detection of Helicobacter pylori is most prominent) are meanwhile well established. However, recent research and development go beyond gastroenterological applications. Sensitive and selective detection of certain volatile organic compounds (VOCs) in a patient's breath, could enable the diagnosis of diseases that are very difficult to diagnose with contemporary techniques. For instance, an appropriate VOC biomarker for early-stage bronchial carcinoma (lung cancer) is n-butane (C₄H₁₀). We present a new optical detection scheme for VOCs that employs an especially compact and simple set-up based on photoacoustic spectroscopy (PAS). This method makes use of the transformation of absorbed modulated radiation into a sound wave. Employing a wavelength-modulated distributed feedback (DFB) diode laser and taking advantage of acoustical resonances of the sample cell, we performed very sensitive and selective measurements on butane. A detection limit for butane in air in the ppb range was achieved. In subsequent research the sensitivity will be successively improved to match the requirements of the medical application. Upon optimization, our photoacoustic sensor has the potential to enable future breath tests for early-stage lung cancer diagnostics.

We have realized a surface plasmon resonance imaging system allowing accurate characterization of biochips. In this paper, the Rouard approach is extended to absorbing layers to model the reflectivity information contained in the multidimensional data. The multidimension potential is also expressed to demonstrate the power of the SPR imaging system. To conclude, towards the development of a biosensor based on SPR, a theoretical study is also performed on the sensitivity to changes in reflectivity of such multidimension optical biosensor. The sensitivity of the sytem shows the power of this biophotonic technology.

The development of reliable methods for the detection of minute amounts of antibodies directly in homogeneous solution represents one of the major tasks in the current research field of molecular diagnostics. We demonstrate the potential of fluorescence correlation spectroscopy (FCS) in combination with quenched peptide-based fluorescence probes for sensitive detection of p53 antibodies directly in homogeneous solution. Single tryptophan (Trp) residues in the sequences of short, synthetic peptide epitopes of the human p53 protein efficiently quench the fluorescence of an oxazine fluorophore attached to the amino terminal ends of the peptides. The fluorescence quenching mechanism is thought to be a photoinduced electron transfer reaction from Trp to the dye enabled by the formation of intramolecular complexes between dye and Trp. Specific recognition of the epitope by the antibody confines the conformational flexibility of the peptide. Consequently, complex formation between dye and Trp is abolished and fluorescence is recovered. Using fluorescence correlation spectroscopy (FCS), antibody binding can be monitored observing two parameters simultaneously: the diffusional mobility of the peptide as well as the quenching amplitude induced by the conformational flexibility of the peptide change significantly upon antibody binding. Our data demonstrate that FCS in combination with fluorescence-quenched peptide epitopes opens new possibilities for the reliable detection of antibody binding events in homogeneous solution.

Background: Arterial bifurcations are commonly the sites of developing atherosclerotic plaque that lead to arterial occlusions and plaque rupture (myocardial infarctions and strokes). Laser induced fluorescence (LIF) spectroscopy provides an effective nondestructive method supplying spectral information on extracellular matrix (ECM) protein composition, specifically collagen and elastin. Purpose: To investigate regional differences in the ECM proteins -- collagen I, III and elastin in unstable plaque by analyzing data from laser-induced fluorescence spectroscopy of human carotid endarterectomy specimens. Methods: Gels of ECM protein extracts (elastin, collagen types I & III) were measured as reference spectra and internal thoracic artery segments (extra tissue from bypass surgery) were used as tissue controls. Arterial segments and the endarterectomy specimens (n=21) were cut into 5mm cross-sectional rings. Ten fluorescence spectra per sampling area were then recorded at 5 sites per ring with argon laser excitation (357nm) with a penetration depth of 200 μm. Spectra were normalized to maximum intensity and analyzed using multiple regression analysis. Tissue rings were fixed in formalin (within 3 hours of surgery), sectioned and stained with H&E or Movat's Pentachrome for histological analysis. Spectroscopy data were correlated with immunohistology (staining for elastin, collagen types I, III and IV). Results: Quantitative fluorescence for the thoracic arteries revealed a dominant elastin component on the luminal side -- confirmed with immunohistology and known artery structure. Carotid endarterectomy specimens by comparison had a significant decrease in elastin signature and increased collagen type I and III. Arterial spectra were markedly different between the thoracic and carotid specimens. There was also a significant elevation (p<0.05) of collagen type I distal to the bifurcation compared to proximal tissue in the carotid specimens. Conclusion: Fluorescence spectroscopy is an effective method for evaluating ECM (collagen and elastin) associated with vascular remodeling despite the considerable variability in the plaque structure. Consistent regional differences were detected in the carotid specimens.

Numerical simulations of time-resolved light transport in inhomogeneous tissues reveal quantitative, 3D-distributions of excitation and fluorescent light. Visualizations generated can assist the optimization of endoscopy-compatible fiber-optic probes and optical imaging systems.

The depth-localization of fluorescent objects having different diameters and embedded within semi-infinite turbid tissue is determined with a model based on the finite element method. The work relies on the time to reach half of the maximum fluorescence intensity. The model is based on a set of two-dependent photon diffusion equations: - the transport of the pulsed laser source (duration 1 ps) and - the transport of the induced fluorescent light excited by the source. The coupling between these equations is due to a source term directly proportional to the scattered fluence rate at the same location. To solve this problem, the method proceeds following the Galerkin formulation added to an implicit finite difference scheme (Backward Euler) to integrate the resulting matrix formulation with respect to time. The meshed domain is axi-symmetrical and takes into account the boundary conditions relative to air-tissue interface. The different computational results show that the fluorescent signals can be used to provide time of flight information about the depth localization of a spherical tumor embedded in a turbid medium. These findings are in good agreement with experimental works.

A reliable theoretical model is essential to reveal information from measured reflectance spectra. Currently, Monte Carlo simulation (MC) is the most popular technique to retrieve optical tissue parameters. However, optical diffusion theory provides an analytic approach that might supersede MC methods due to faster, more efficient algorithms. Diffuse skin reflectance in the 400-800 nm wavelength range was simulated by Monte Carlo and diffusion theory. The impact of detection geometry and source distribution was investigated, and experimental data from bruised and normal skin were fitted using a three layer diffusion model. Spectra from diffusion theory were within 5% of the MC results, and the fit between the two methods was further improved by scaling the dermal absorption parameters with a constant factor. The measurement geometry was found to be of minor impact. Diffusion theory was found to have wide applicability due to fast, efficient algorithms that allow efficient evaluation of experimental data.

The potentialities offered by time-gated transillumination of inhomogeneous tissue slab media are numerically investigated in this paper. A finite element model is firstly presented in order to solve the time-dependent light transport in mutiple-scattering optical media containing different embedded objects. The numerical procedure is based on the Galerkin formulation added to an implicite finite difference scheme (Backward Euler) to integrate the resulting matrix formulation with respect to time. The meshed domain refers to Cartesian-coordinates system (x,y) such that the computational grid can be adapted to scan along the longitudinal coordinate x. It takes into account the boundary conditions relative to air-tissue interfaces. The use of the method is demonstrated by the forward computations of time-gated intensities, resulting from line scans across either partially absorbing or scattering cylindrical objects. The overall computations confirm that time-gating technique is very sensitive to local variations in optical properties that are due to hidden objects in turbid media. It is also shown that the lateral localization of these inclusions is enhanced when the time-gate width (Δt) is decreased to about 30 ps.

Optimization of device-tissue interface parameters may lead to an improvement in the efficacy of fluorescence spectroscopy for minimally-invasive disease detection. Although illumination-collection geometry has been shown to have a strong influence on the spatial origin of detected fluorescence, the performance of devices which deliver and/or collect light at oblique incidence are not well characterized or understood. Simulations were performed using a Monte Carlo model of light propagation in homogeneous tissue in order to identify and describe general trends in the intensity and spatial origin of fluorescence detected by angled geometries. Specifically, the influence of illumination angle, collection angle and illumination-collection spot separation distance were investigated for low and high attenuation tissue cases. Results indicated that oblique-incidence geometries have the potential to enhance the selective interrogation of superficial or subsurface fluorophores at user-selectable depths up to about 0.5 mm. Detected fluorescence intensity was shown to increase significantly with illumination and collection angle. Improved selectivity and signal intensity over normal-incidence geometries resulted from the overlap of illumination and collection cones within the tissue. Cases involving highly attenuating tissue produced a moderate reduction in the depth of signal origin. While Monte Carlo modeling indicates that oblique-incidence designs can facilitate depth-selective fluorescence spectroscopy, optimization of device performance will require application-specific consideration of optical and biological parameters.

Age-related macular degeneration (AMD) is the most frequent cause for blindness for person older than 65 years in western countries. Besides the subjective pain, it is also an economic problem in the ageing population. As the pathomechanism is unknown, no curative treatment is possible. International research for discovering of early age-related alterations at the fundus is directed on in vivo measurements of autofluorescence. One way is the measurement of fluorescence spectra. Unfortunately, any selective excitation of fluorophores is excluded by the absorption edge of the ocular media at 400 nm. Furthermore, the shape of fluorescence spectra is influenced by the spectral absorption of layers in front of the emitting fluorophore. Weakly emitting fluorophores are covered by intensive emitting substances. The most serious limitation in fluorescence measurements of the living human fundus is the maximal permissible exposure. For that reason, fluorescence spectra of the fundus can not be detected with a high spatial resolution. The detection of dynamic fluorescence provides substance-specific lifetimes Ti, amplitudes A_i, and information about the relative contribution Q_i of components in fluorescence images. As these parameters are calculated for each image point, diagrams of T_i vs. T_j, A_i vs. A_j, and Q_i vs. Q_j can be drawn, in which specific clusters appear for healthy subjects or AMD - patients. The projection of lifetime - clusters onto corresponding axis represents histogram of the considered lifetime. The slope in the correlation between Ai and Aj can also be used as a discriminating mark. Considering image lines as intersection through characteristic anatomical structures (optic disc, macula) profiles of T_i, A_i, or Q_i show changes of these parameters (e.g. depigmentation) as function of location, which might be specific for AMD.

We present a multi-dimensional TCSPC technique that simultaneously records the photon distribution over the time in the fluorescence decay, the wavelength, and the coordinates of a two-dimensional scan. We demonstrate the application of the technique to single-point autofluorescence measurements of skin, to multi-spectral fluorescence lifetime microscopy, and ophthalmic imaging.

Tissue contains many natural fluorophores and therefore by exploiting autofluorescence, we can obtain information from tissue with less interference than conventional histological techniques. However, conventional intensity imaging is prone to artifacts since it is an absolute measurement. Fluorescence lifetime and spectral measurements are relative measurements and therefore allow for better measurements. We have applied FLIM and hyperspectral FLIM to the study of articular cartilage and its disease arthritis. We have analyzed normal human articular cartilage and cartilage which was in the early stages of disease. In this case, it was found that FLIM was able to detect changes in the diseased tissue that were not detectable with the conventional diagnosis. Specifically, the fluorescence lifetimes (FL) of the cells were different between the two samples. We have also applied hyperspectral FLIM to degraded cartilage through treatment with interleukin-1. In this case, it was found that there was a shift in the emission spectrum with treatment and that the lifetime had also increased. We also showed that there was greater contrast between the cells and the extracellular matrix (ECM) at longer wavelengths.

Urinary cytology is employed in diagnostic guidelines of bladder cancer in anatomo-pathological laboratories mostly for its ability to diagnose non detectable cancers using cystoscopy, but also because it is a non-invasive and non-constraining technique for a regular follow-up of the more exposed populations. The impossibility to detect such cancers is mainly due to their localization either in the bladder or in the upper urinary tract and the prostate. However, urinary cytology lacks sensitivity, especially for the detection of low grade low stage tumors due to inherent limitation of morphological criteria to distinguish low grade tumor cells from normal urothelial cells. For this purpose, we developed, in addition to urinary cytology, an original screening of these cytological slides by using spectrally-resolved and time-resolved fluorescence as a contrast factor, without changing any parameters in the cytological slide preparation. This method takes advantage of a femtosecond Ti:sapphire laser, continuously tunable in the spectral range 700-950 nm allowing the observation of most endogenous cellular chromophores by biphotonic excitation. A commercial confocal microscope was also used in the measurements allowing an excitation of the samples between 458 nm and 633 nm. We observed that the fluorescence emission is differentially distributed in normal and pathological urothelial cells. Spectral- and time-resolved measurements attested this difference over about one hundred cases which have been tested to confirm the high accuracy of this non-invasive technique.

Photobiological research in the last decades has shown the existence of Delayed Luminescence in biological tissue, which presents an excitation spectrum with a peak within the UVA region and can be detected with sophisticated photomultiplier systems. Based on these findings, a new and powerful tool able to measure the UV-A-laser-induced Delayed Luminescence emission of cultured cells was developed, with the intention to detect biophysical changes between carcinogenic and normal cells. Indeed noticeable differences have been found in the time resolved emission spectrum of delayed luminescence of cell cultures of human fibroblast and human melanoma. This new, powerful and non-invasive technique, in principle, could be applied in all fields of skin research, such as the investigation of skin abnormalities and to test the effect of products involved in regeneration, anti-aging and UV-light protection in order to prevent skin cancer.

Decays of tyrosine fluorescence in protein-ligand complexes are described by a model of continuous distribution of fluorescence lifetimes. Resulted analytical power-like decay function provides good fits to highly complex fluorescence kinetics. Moreover, this is a manifestation of so-called Tsallis q-exponential function, which is suitable for description of the systems with long-range interactions, memory effect, as well as with fluctuations of the characteristic lifetime of fluorescence. The proposed decay functions were applied to analysis of fluorescence decays of tyrosine in a protein, i.e. the enzyme purine nucleoside phosphorylase from E. coli (the product of the deoD gene), free in aqueous solution and in a complex with formycin A (an inhibitor) and orthophosphate (a co-substrate). The power-like function provides new information about enzyme-ligand complex formation based on the physically justified heterogeneity parameter directly related to the lifetime distribution. A measure of the heterogeneity parameter in the enzyme systems is provided by a variance of fluorescence lifetime distribution. The possible number of deactivation channels and excited state mean lifetime can be easily derived without a priori knowledge of the complexity of studied system. Moreover, proposed model is simpler then traditional multi-exponential one, and better describes heterogeneous nature of studied systems.

The glucose concentration in arterial plasma has immediate effects on the optical properties of blood-bearing tissue due primarily to the alteration of refractive index mismatch between the scattering particles (red blood cells) and the medium (plasma). The influence of these effects on pulse oximetry is investigated using a numerical model based on Mie theory. The objective is to determine whether or not physiological fluctuations in blood glucose levels could sufficiently vary the optical properties to shift the calibration curve of a commercial pulse oximeter significantly.

Theoretical considerations concerning new method for interstitial photodynamic therapy light dosimetry and distribution at different depths, by means of single-fiber multi-decay-probes is presented, basing on decomposition of multi-exponential decay. Successful decomposition of up to 25 and up to 7 components of the modelled fiber probe was achieved with noiseless and noise containing decay curves respectively. The impact of noise and distribution of excitation intensity were examined and optimisation algorithm was briefly described.

The transmittance photoplethysmographic signals recorded with multiple NIR laser diodes in athletes along a maximal exercise test by treadmill ergometer and the results after processing are presented in comparison to the established reference techniques.

The scope of this work was to determine the Kubelka-Munk scattering and absorption coefficients of healthy and atherosclerotic animal model aorta, from 200 to 1100 nm. Furthermore, using the measured and calculated optical properties, special algorithms were developed in order to discriminate healthy from diseased aorta. Diffuse reflectance and total transmittance were measured via a dual beam diffuse reflectance spectrometer. Inverse Kubelka-Munk Model was applied to calculate the diffusion scattering and absorption coefficients. Diffuse absorption coefficients varied from ~200 cm^-1 at 300 nm to ~3 cm^-1 at 1100 nm. Kubelka-Munk scattering coefficients ranged from ~100 cm^-1 at 200 nm to ~6 cm^-1 at 1100 nm. Appropriate discrimination algorithms were developed in order to characterize a specimen as healthy or atherosclerotic. The first algorithm was based on the ratio of diffuse reflectance to the reflectance in infinity. The gradient of this ratio at 390 and 440 nm effectively separated healthy from atherosclerotic aorta. The discrimination between the two groups was succeeded using multivariate statistical analysis and verified by histopathology. The second discrimination algorithm was based on the ratio of diffuse reflectance to the baseline reflectance of each specimen. Effective discrimination of healthy and atherosclerotic aorta was achieved at 370 and 500 nm.

In this paper the real and imaginary parts of the refractive index of hemoglobin solutions were determined in the wavelength range of 250 to 1100 nm using Fresnel reflectance and transmittance measurements. The determined real parts of the refractive indices are on average 0.02 units higher than the values found in the literature.

Glucose content monitoring is of great importance today due to a number of people suffering from diabetes. In this paper, laser pulses propagation in a sample of aqueous Intralipid solution with glucose is simulated by Monte Carlo method. Effect of glucose is based on refractive index matching of Intralipid vesicles and surrounding water if glucose is added. Temporal profiles of femtosecond pulses (906 nm) diffusely scattered within a 2-mm thick plain glass cuvette with a skin phantom are registered in backward direction by two fiber-optics detectors 0.30 mm in diameter with numerical apertures of 0.19, 0.29, and 0.39. It is revealed that glucose content within the physiological range (100-500 mg/dl) can be detected because of the effect of glucose on the peak pulse intensity and on the area under the pulse temporal profile (energy of the registered pulse).