This PDF file contains the front matter associated with SPIE Proceedings Volume 9318, including the Title Page, Copyright information, Table of Contents, Introduction (if any), and Conference Committee listing.

Diabetic foot ulcer is the most devastating complication of diabetes that is still un-recognized. The treatment costs of these ulcers are very high to eventually save the leg/foot from amputation. To date, clinicians employ visual inspection of the wound site during its standard 4-week of healing process via monitoring of surface granulation. There is a need to develop on-site, low-cost imaging tools that can monitor the wound healing process periodically during the standard 4-week treatment process. A novel ultra-portable near-infrared optical scanner (NIROS) has been developed at the Optical Imaging Laboratory that can perform non-contact 2D area imaging of the wound site. Non-contact optical imaging studies were carried on diabetic subjects with foot ulcers (at Somesh Diabetic Foot Clinic, India) that were of healing and non-healing nature. A 710 nm LED source and a compact NIR sensitive camera were employed during non-contact imaging of the diabetic foot in order to obtain the near-infrared absorption images. From these preliminary studies it was observed that the non-healing wounds had a greater absorption contrast with respect to the normal site, unlike in the healing wounds. Demonstrating the ability of NIROS to differentiate healing vs. non-healing wounds in diabetic subjects can potentially impact early intervention in the treatment of diabetic foot ulcers.

Histologic grade is a very important, but underappreciated, parameter of breast cancer aggressiveness. Despite its importance, it has historically not been included as one of the criteria for staging of this cancer. In this study, spectral fluorescence profiles from patients with breast carcinoma were acquired. Ratios of emission peaks at 340 over 440,460 nm from biomolecules in malignant and normal samples were calculated. Cancerous over normal ratios (double ratio (DR) method) were evaluated with respect to tumor characteristics. Increased tryptophan content in breast cancer tissues correlates strongly with high grade, but not with lymph node metastases, estrogen receptor, progesterone receptor or Her-2-Neu receptor status.

Light scattering and transmission of optical Laguerre Gaussian (LG) vortex beams with different orbital angular momentum (OAM) states in turbid scattering media were investigated in comparison with Gaussian (G) beam. The scattering media used in the experiments consist of various sizes and concentrations of latex beads in water solutions. The LG beams were generated using a spatial light modulator in reflection mode. The ballistic transmissions of LG and G beams were measured with different ratios of thickness of samples (z) to scattering mean free path (l_s) of the turbid media, z/l_s. The results show that in the ballistic region where z/ls is small, the LG and G beams show no significant difference, while in the diffusive region where z/l_s is large, LG beams show higher transmission than Gaussian beam. In the diffusive region, the LG beams with higher orbital angular momentum L values show higher transmission than the beams with lower L values. The transition points from ballistic to diffusive regions for different scattering media were studied and determined.

Photoacoustic (PA) measurements encode the information associated with both physical microstructures and chemical contents in biological tissues. A two-dimensional physio-chemical spectrogram (PCS) can be formulated by combining the power spectra of PA signals acquired at a series of optical wavelengths. The analysis of PCS, or namely PA physio-chemical analysis (PAPCA), enables the quantification of the concentrations and the spatial distributions of a variety of chemical components in the tissue. The chemical components and their distribution are the two major features observed in the biopsy procedures which have been regarded as the gold standard of the diagnosis of many diseases. Taking non-alcoholic fatty liver disease and prostate cancer for example, this study investigates the feasibility of PAPCA in characterizing the histopathological changes in the diseased conditions in biological tissue. A catheter based setup facilitating measurement in deep tissues was also proposed and tested.

Optical coherence tomography (OCT) provides significant advantages of high-resolution (approaching the histopathology level) real-time imaging of tissues without use of contrast agents. Based on these advantages, the microstructural features of tumors can be visualized and detected intra-operatively. However, it is still not clinically accepted for tumor margin delineation due to poor specificity and accuracy. In contrast, Raman spectroscopy (RS) can obtain tissue information at the molecular level, but does not provide real-time imaging capability. Therefore, combining OCT and RS could provide synergy. To this end, we present a tissue analysis and classification method using both the slope of OCT intensity signal versus depth and the principle components from the RS spectrum as the indicators for tissue characterization. Our pilot experiments were performed on mouse kidneys, livers, and small intestines. The prediction accuracy with five-fold cross validation of the method has been evaluated by support vector machine method. The results demonstrate that RS can effectively improve tissue classification compared to OCT alone. Next, we demonstrate that the boundary between myxoid liposarcoma and normal fat which is easily identifiable both Raman and OCT. In cases where structural images are indistinguishable, for example, in normal fat and well differentiated liposarcoma (WDLS) or gastrointestinal sarcoma tumor (GIST) and Myxoma, distinct molecular spectra have been obtained. The results suggest RS can effectively complement OCT to tumor boundary demarcation with high specificity.

Two different optical fiber probes for combined Raman and fluorescence spectroscopic measurements were designed, developed and used for tissue diagnostics. Two visible laser diodes were used for fluorescence spectroscopy, whereas a laser diode emitting in the NIR was used for Raman spectroscopy. The two probes were based on fiber bundles with a central multimode optical fiber, used for delivering light to the tissue, and 24 surrounding optical fibers for signal collection. Both fluorescence and Raman spectra were acquired using the same detection unit, based on a cooled CCD camera, connected to a spectrograph. The two probes were successfully employed for diagnostic purposes on various tissues in a good agreement with common routine histology. This study included skin, brain and bladder tissues and in particular the classification of: malignant melanoma against melanocytic lesions and healthy skin; urothelial carcinoma against healthy bladder mucosa; brain tumor against dysplastic brain tissue. The diagnostic capabilities were determined using a cross-validation method with a leave-one-out approach, finding very high sensitivity and specificity for all the examined tissues. The obtained results demonstrated that the multimodal approach is crucial for improving diagnostic capabilities. The system presented here can improve diagnostic capabilities on a broad range of tissues and has the potential of being used for endoscopic inspections in the near future.

During breast conserving surgery (BCS), which is the preferred approach to treat most early stage breast cancers, the surgeon attempts to excise the tumor volume, surrounded by thin margin of normal tissue. The intra-operative assessment of cancerous areas is a challenging procedure, with the surgeon usually relying on visual or tactile guidance. This study evaluates whether time-resolved fluorescence spectroscopy (TRFS) presents the potential to address this problem. Point TRFS measurements were obtained from 19 fresh tissue slices (7 patients) and parameters that characterize the transient signals were quantified via constrained least squares deconvolution scheme. Fibrotic tissue (FT, n=69), adipose tissue (AT, n=76), and invasive ductal carcinoma (IDC, n=27) were identified in histology and univariate statistical analysis, followed by multi-comparison test, was applied to the corresponding lifetime data. Significant differentiation between the three tissue types exists at 390 nm and 500 nm bands. The average lifetime is 3.23±0.74 ns for AT, 4.21±0.83 ns for FT and 4.71±0.35 ns (p<0.05) for IDC at 390 nm. Due to the smaller contribution of collagen in AT the average lifetime value is different from FT and IDC. Additionally, although intensity measurements do not show difference between FT and IDC, lifetime can distinguish them. Similarly, in 500 nm these values are 7.01±1.08 ns, 5.43±1.05 ns and 4.39±0.88 ns correspondingly (p<0.05) and this contrast is due to differentiation in retinol or flavins relative concentration, mostly contributing to AT. Results demonstrate the potential of TRFS to intra-operatively characterize BCS breast excised tissue in real-time and assess tumor margins.

Cancer tissue is frequently impossible to distinguish from normal brain during surgery. Gliomas are a class of brain cancer which invade into the normal brain. If left unresected, these invasive cancer cells are the source of glioma recurrence. Moreover, these invasion areas do not show up on standard-of-care pre-operative Magnetic Resonance Imaging (MRI). This inability to fully visualize invasive brain cancers results in subtotal surgical resections, negatively impacting patient survival. To address this issue, we have demonstrated the efficacy of single-point in vivo Raman spectroscopy using a contact hand-held fiber optic probe for rapid detection of cancer invasion in 8 patients with low and high grade gliomas. Using a supervised machine learning algorithm to analyze the Raman spectra obtained in vivo, we were able to distinguish normal brain from the presence of cancer cells with sensitivity and specificity greater than 90%. Moreover, by correlating these results with pre-operative MRI we demonstrate the ability to detect low density cancer invasion up to 1.5cm beyond the cancer extent visible using MRI. This represents the potential for significant improvements in progression-free and overall patient survival, by identifying previously undetectable residual cancer cell populations and preventing the resection of normal brain tissue. While the importance of maximizing the volume of tumor resection is important for all grades of gliomas, the impact for low grade gliomas can be dramatic because surgery can even be curative. This convenient technology can rapidly classify cancer invasion in real-time, making it ideal for intraoperative use in brain tumor resection.

A novel microscopy method that takes advantage of shallow photon penetration using ultraviolet-range excitation and exogenous fluorescent stains is described. This approach exploits the intrinsic optical sectioning function when exciting tissue fluorescence from superficial layers to generate images similar to those obtainable from a physically thinsectioned tissue specimen. UV light in the spectral range from roughly 240-275 nm penetrates only a few microns into the surface of biological specimens, thus eliminating out-of-focus signals that would otherwise arise from deeper tissue layers. Furthermore, UV excitation can be used to simultaneously excite fluorophores emitting across a wide spectral range. The sectioning property of the UV light (as opposed to more conventional illumination in the visible range) removes the need for physical or more elaborate optical sectioning approaches, such as confocal, nonlinear or coherent tomographic methods, to generate acceptable axial resolution. Using a tunable laser, we investigated the effect of excitation wavelength in the 230-350 nm spectral range on excitation depth. The results reveal an optimal wavelength range and suggest that this method can be a fast and reliable approach for rapid imaging of tissue specimens. Some of this range is addressable by currently available and relatively inexpensive LED light sources. MUSE may prove to be a good alternative to conventional, time-consuming, histopathology procedures.

Imaging of vessel structures can be useful for investigation of endothelial function, angiogenesis and hyper-vascularization. This can be challenging for hyperspectral tissue imaging due to photon scattering and absorption in other parts of the tissue. Real-time processing techniques for enhancement of vessel contrast in hyperspectral tissue images were investigated. Wavelet processing and an inverse diffusion model were employed, and compared to band ratio metrics and statistical methods. A multiscale vesselness filter was applied for further enhancement. The results show that vessel structures in hyperspectral images can be enhanced and characterized using a combination of statistical, numerical and more physics informed models.

Hyperspectral imaging (HSI) systems have the potential to combine morphological and spectral information to provide detailed and high sensitivity readouts in biological and medical applications. As HSI enables simultaneous detection in several spectral bands, the technology has significant potential for use in real-time multiplexed contrast agent studies. Examples include tumor detection in intraoperative and endoscopic imaging as well as histopathology. A multiplexed readout from multiple disease targets, such as cell surface receptors overexpressed in cancer cells, could improve both sensitivity and specificity of tumor identification. Here, we evaluate a commercial, compact, near-infrared HSI sensor that has the potential to enable low cost, video rate HSI for multiplexed fluorescent contrast agent studies in biomedical applications. The hyperspectral imager, based on a monolithically integrated Fabry-Perot etalon, has 70 spectral bands between 600-900 nm, making it ideal for this application. Initial calibration of the imager was performed to determine wavelength band response, quantum efficiency and the effect of F-number on the spectral response. A platform for wide-field fluorescence imaging in reflectance using fluorophore specific LED excitation was then developed. The applicability of the imaging platform for simultaneous readout of multiple fluorophore signals was demonstrated using a dilution series of Alexa Fluor 594 and Alexa Fluor 647, showing that nanomolar fluorophore concentrations can be detected. Our results show that the HSI system can clearly resolve the emission spectra of the two fluorophores in mixtures of concentrations across several orders of magnitude, indicating a high dynamic range performance. We therefore conclude that the HSI sensor tested here is suitable for detecting fluorescence in biomedical imaging applications.

Oral cancers are considered to be one of the most commonly occurring malignancy worldwide. Over 70% of the cases report to the doctor only in advanced stages of the disease, resulting in poor survival rates. Hence it is necessary to detect the disease at the earliest which may increase the five year survival rate up to 90%. Among various optical spectroscopic techniques, Raman spectroscopy has been emerged as a tool in identifying several diseased conditions, including oral cancers. Around 30 - 80% of the malignancies of the oral cavity arise from premalignant lesions. Hence, understanding the molecular/spectral differences at the premalignant stage may help in identifying the cancer at the earliest and increase patient’s survival rate. Among various bio-fluids such as blood, urine and saliva, urine is considered as one of the diagnostically potential bio-fluids, as it has many metabolites. The distribution and the physiochemical properties of the urinary metabolites may vary due to the changes associated with the pathologic conditions. The present study is aimed to characterize the urine of 70 healthy subjects and 51 pre-malignant patients using Raman spectroscopy under 785nm excitation, to know the molecular/spectral differences between healthy subjects and premalignant conditions of oral malignancy. Principal component analysis based Linear discriminant analysis were also made to find the statistical significance and the present technique yields the sensitivity and specificity of 86.3% and 92.9% with an overall accuracy of 90.9% in the discrimination of premalignant conditions from healthy subjects urine.

Confocal Raman microspectroscopy allows in-depth molecular and conformational characterization of biological tissues non-invasively. Unfortunately, spectral distortions occur due to elastic scattering. Our objective is to correct the attenuation of in-depth Raman peaks intensity by considering this phenomenon, enabling thus quantitative diagnosis. In this purpose, we developed PDMS phantoms mimicking skin optical properties used as tools for instrument calibration and data processing method validation. An optical system based on a fibers bundle has been previously developed for in vivo skin characterization with Diffuse Reflectance Spectroscopy (DRS). Used on our phantoms, this technique allows checking their optical properties: the targeted ones were retrieved. Raman microspectroscopy was performed using a commercial confocal microscope. Depth profiles were constructed from integrated intensity of some specific PDMS Raman vibrations. Acquired on monolayer phantoms, they display a decline which is increasing with the scattering coefficient. Furthermore, when acquiring Raman spectra on multilayered phantoms, the signal attenuation through each single layer is directly dependent on its own scattering property. Therefore, determining the optical properties of any biological sample, obtained with DRS for example, is crucial to correct properly Raman depth profiles. A model, inspired from S.L. Jacques's expression for Confocal Reflectance Microscopy and modified at some points, is proposed and tested to fit the depth profiles obtained on the phantoms as function of the reduced scattering coefficient. Consequently, once the optical properties of a biological sample are known, the intensity of deep Raman spectra distorted by elastic scattering can be corrected with our reliable model, permitting thus to consider quantitative studies for purposes of characterization or diagnosis.

This paper describes a novel image processing method for endoscopy that enhances the appearance of microstructures on mucosa. The new technique employs two pairs of parallel- and crossed-nicols polarimetric images, from which an averaged subtracted polarization image (AVSPI) is calculated. Experiments were first executed using a manual experimental setup with ring-type lighting, two rotating polarizers and a color camera. A new objective evaluation method that uses texture analysis (GLCM) was developed and applied to evaluation of the enhanced microstructure images. Experiments using excised porcine stomach tissue showed better results than with conventional color intensity image processing. Next, an online rigid-type polarimetric endoscope system using a polarized ring-shaped LED and a special three-CCD color polarimetric camera was developed. The two types of equipment described above are quite different as to extinction ratio values, but show similarly enhanced image quality. Our results show that polarimetric endoscopy is not only effective but also practical for hardware implementation.

Reflectance spectroscopy contains information of scatterers and absorbers present inside biological tissues and has been successfully used to diagnose disease. Success of any diagnostic tool depends upon the potential of statistical algorithm to extract appropriate diagnostic features from the measured optical data. In our recent study, we have used the potential of the classification algorithm, Nonlinear Maximum Representation and Discrimination Features (NMRDF) to extract important diagnostic features from reflectance spectra of normal and dysplastic human cervical tissue. This NMRDF algorithm uses the higher order correlation information in the input data, which helps to represent the asymmetrically distributed data and provides the closed form solution of the nonlinear transform for maximum discrimination. We have recorded unpolarized, co and cross-polarized reflectance spectra from 350nm to 650nm, illuminating the human cervical tissue epithelium with white light source. A total of 139 samples were divided into training and validation data sets. The input parameters were optimized using training data sets to extract the appropriate nonlinear features from the input reflectance spectra. These extracted nonlinear features are used as input for nearest mean classifier to calculate the sensitivity and specificity for both training as well as validation data sets. We have observed that co-polarized components provide maximum sensitivity and specificity compared to cross-polarized components and unpolarized data. This is expected since co-polarized light provides subsurface information while cross-polarized and unpolarized data mask the vital epithelial information through high diffuse scattering.

Optical spectroscopy has been considered a promising method for cancer detection for past thirty years because of its advantages over the conventional diagnostic methods of no tissue removal, minimal invasiveness, rapid diagnoses, less time consumption and reproducibility since the first use in 1984. It offers a new armamentarium. Human tissue is mainly composed of extracellular matrix of collagen fiber, proteins, fat, water, and epithelial cells with key molecules in different structures. Tissues contain a number of key fingerprint native endogenous fluorophore molecules, such as tryptophan, collagen, elastin, reduced nicotinamide adenine dinucleotide (NADH), flavin adenine dinucleotide (FAD) and porphyrins. It is well known that abnormalities in metabolic activity precede the onset of a lot of main diseases: carcinoma, diabetes mellitus, atherosclerosis, Alzheimer, and Parkinson's disease, etc. Optical spectroscopy may help in detecting various disorders. Conceivably the biochemical or morphologic changes that cause the spectra variations would appear earlier than the histological aberration. Therefore, “optical biopsy” holds a great promise as clinical tool for diagnosing early stage of carcinomas and other deceases by combining with available photonic technology (e.g. optical fibers, photon detectors, spectrographs spectroscopic ratiometer, fiber-optic endomicroscope and nasopharyngoscope) for in vivo use. This paper focuses on various methods available to detect spectroscopic changes in tissues, for example to distinguish cancerous prostate tissues and/or cells from normal prostate tissues and/or cells. The methods to be described are fluorescence, stokes shift, scattering, Raman, and time-resolved spectroscopy will be reviewed. The underlying physical and biological basis for these optical approaches will be discussed with examples. The idea is to present some of the salient works to show the usefulness and methods of Optical Biopsy for cancer detection and show new directions.

Native fluorescence spectroscopy offers an important role in cancer discrimination. It is widely acknowledged that the emission spectrum of tissue is a superposition of spectra of various salient fluorophores. In this study, the native fluorescence spectra of human cancerous and normal breast tissues excited by selected wavelength of 300 nm are used to investigate the key building block fluorophores: tryptophan and reduced nicotinamide adenine dinucleotide (NADH). The basis spectra of these key fluorophores' contribution to the tissue emission spectra are obtained by nonnegative constraint analysis. The emission spectra of human cancerous and normal tissue samples are projected onto the fluorophore spectral subspace. Since previous studies indicate that tryptophan and NADH are key fluorophores related with tumor evolution, it is essential to obtain their information from tissue fluorescence but discard the redundancy. To evaluate the efficacy of for cancer detection, linear discriminant analysis (LDA) classifier is used to evaluate the sensitivity, and specificity. This research demonstrates that the native fluorescence spectroscopy measurements are effective to detect changes of fluorophores' compositions in tissues due to the development of cancer.

Recent studies have demonstrated the potential advantages of the use of Raman spectroscopy in the biomedical field due to its rapidity and noninvasive nature. In this study, Raman spectroscopy is applied as a method for differentiating between bacteria isolates for Gram status and Genus species. We created models for identifying 28 bacterial isolates using spectra collected with a 785 nm laser excitation Raman spectroscopic system. In order to investigate the groupings of these samples, partial least squares discriminant analysis (PLSDA) and hierarchical cluster analysis (HCA) was implemented. In addition, cluster analyses of the isolates were performed using various data types consisting of, biochemical tests, gene sequence alignment, high resolution melt (HRM) analysis and antimicrobial susceptibility tests of minimum inhibitory concentration (MIC) and degree of antimicrobial resistance (SIR). In order to evaluate the ability of these models to correctly classify bacterial isolates using solely Raman spectroscopic data, a set of 14 validation samples were tested using the PLSDA models and consequently the HCA models. External cluster evaluation criteria of purity and Rand index were calculated at different taxonomic levels to compare the performance of clustering using Raman spectra as well as the other datasets. Results showed that Raman spectra performed comparably, and in some cases better than, the other data types with Rand index and purity values up to 0.933 and 0.947, respectively. This study clearly demonstrates that the discrimination of bacterial species using Raman spectroscopic data and hierarchical cluster analysis is possible and has the potential to be a powerful point-of-care tool in clinical settings.

The objective of this study was to find out the emission spectral fingerprints for discrimination of human colorectal and gastric cancer from normal tissue in vitro by applying native fluorescence. The native fluorescence (NFL) and Stokes shift spectra of seventy-two human cancerous and normal colorectal (colon, rectum) and gastric tissues were analyzed using three selected excitation wavelengths (e.g. 300 nm, 320 nm and 340 nm). Three distinct biomarkers, tryptophan, collagen and reduced nicotinamide adenine dinucleotide hydrate (NADH), were found in the samples of cancerous and normal tissues from eighteen subjects. The spectral profiles of tryptophan exhibited a sharp peak in cancerous colon tissues under a 300 nm excitation when compared with normal tissues. The changes in compositions of tryptophan, collagen, and NADH were found between colon cancer and normal tissues under an excitation of 300 nm by the non-negative basic biochemical component analysis (BBCA) model.

Fluorescence profiles from breast cancer and breast normal tissue samples with excitation wavelengths at 280 nm and 340 nm were obtained using the conventional LS-50 Perkin-Elmer spectrometer. Fluorescence ratios from these tissue samples, demonstrated by emission peaks at 340 nm, 440 nm and 460 nm and likely representing tryptophan and NADH, show increased relative content of tryptophan in malignant samples. Double ratio (DR) techniques were used to measure the severity of disease. The single excitation double ratio (Single-DR) method utilizes the emission intensity peaks from the spectrum acquired using a single excitation of 280 nm; while the dual excitation double ratio (dual-DR) method utilizes the emission intensity peaks from the spectra acquired using an excitation of 280 nm and 340 nm. Single-DR and dual-DR from 13 patients with breast carcinoma were compared in terms of their efficiency to distinguish high from low/intermediate tumors. Similar results were found with both methods. Results suggest that dual excitation wavelengths may be as effective as single excitation wavelength in calculating the relative content of biomolecules in breast cancer tissue, as well as for the assessment of the malignant potential of these tumors.

Raman spectroscopy has been a powerful tool in various fields of science and technology ranging from analytical chemistry to biomedical imaging. In spite of unique features, Raman spectroscopy has also some limitations. Among them are weak Raman signal compared to strong fluorescence and relatively complicated setup with expensive and bulky spectrometer. In order to increase the sensitivity of Raman technique, many clever attempts have been made and some of them were very successful including CARS, SRS, and so on. However, these still requires expensive and more complicated setup. In this work, we have attempted to build a real-time compact Raman sensor without spectrometer. Conventional spectrometer was replaced with a narrow-band optical filter and alternatively modulated two lasers with slightly different wavelengths. At one laser, Raman signal from a target molecule was transmitted through the optical filter. At the other laser, this signal was blocked by the optical filter and could not be detected by photon detector. The alternative modulation of two lasers will modulate the Raman signal from a target molecule at the same modulation frequency. This modulated weak Raman signal was amplified by a lock-in amplifier. The advantages of this setup include compactness, low cost, real-time monitoring, and so on. We have tested the sensitivity of this setup and we found that it doesn’t have enough sensitivity to detect single molecule-level, but it is still good enough to monitor the change of major chemical composition in the sample.

RR spectra of brain normal tissue, gliomas in low grade I and II, and malignant glioma tumors in grade III and IV were measured using a confocal micro Raman spectrometer. This report focus on the relative contents of tryptophan (W) in various grades of brain glioma tumors by the intrinsic molecular resonance Raman (RR) spectroscopy method using the 1588cm^-1 of tryptophan mode by 532 nm excitation. The RR spectra of key fingerprints of tryptophan, with a main vibrational mode at 1588cm^-1 (W8b), were observed. It was found that tryptophan contribution was accumulated in grade I to IV gliomas and the mode of 1588cm^-1 in grade III and IV malignant gliomas were enhanced by resonance.