This PDF file contains the front matter associated with SPIE Proceedings volume 7171, including the Title Page, Copyright information, Table of Contents, Introduction, and the Conference Committee listing.

Image-guided near infrared spectroscopy (IG-NIRS) can provide high-resolution vascular, metabolic and molecular characterization of localized tissue volumes in-vivo. The approach for IG-NIRS uses hybrid systems where the spatial anatomical structure of tissue obtained from standard imaging modalities (such as MRI) is combined with tissue information from diffuse optical imaging spectroscopy. There is need to optimize these hybrid systems for large-scale clinical trials anticipated in the near future in order to evaluate the feasibility of this technology across a larger population. However, existing computational methods such as the finite element method mesh arbitrary image volumes, which inhibit automation, especially with large numbers of datasets. Circumventing this issue, a boundary element method (BEM) for IG-NIRS systems in 3-D is presented here using only surface rendering and discretization. The process of surface creation and meshing is faster, more reliable, and is easily generated automatically as compared to full volume meshing. The proposed method has been implemented here for multi-spectral non-invasive characterization of tissue. In phantom experiments, 3-D spectral BEM-based spectroscopy recovered the oxygen dissociation curve with mean error of 6.6% and tracked variation in total hemoglobin linearly.

Spectrally-resolved BioLuminescence optical Tomography (BLT) aims to reconstruct tomographic images of Luciferase activity within a volume using spectrally resolved emission data from internal bioluminescence sources. Uniqueness as associated with non-spectrally resolved intensity-based BioLuminescence Tomography is an established drawback and it is also demonstrated that without the use of a non-negative constraint inverse algorithm the solution is inaccurate. Reconstructed images of bioluminescence are presented showing that the location of the internal bioluminescence source can be obtained with 0.5 mm accuracy. The concept of the reciprocity approach for image reconstruction is outlined providing a dramatic improvement in computational time without loss to image accuracy.

Magnetic resonance (MR) guided diffuse optical spectroscopy (DOS) has shown promise in several case studies in aiding the characterization of breast lesions[1, 2]. It has been proposed that the increased quantification and resolution with a priori structural guidance yields higher diagnostic value in characterizing tumors. To date, these systems have merged MR anatomical recovery with optical contrast recovery. However, the MR has a wealth of spectral and functional data that may aid in further improving lesion characterization by appending both new and overlapping physiological information to optical methods. It has been well documented that spectral recovery of water and lipids is inaccurate with few wavelengths. Yet, recovery of these chromophores is important both because of the possible importance of these as indicators of breast cancer, adema, and inflammation. In addition, crosstalk between water and oxyhemoglobin may lead to erroneous tissue properties, which may affect lesion diagnosis. The use of multiple MR sequences with DOS enables the separation of water and lipids via MRI, and improves recovery of tissue oxygenation and hemoglobin content. However, in most cases, MRI is not a quantitative device; this paper investigates the best reconstruction methods to incorporate this data into the optical reconstruction for quantitatively accurate chromophore recovery in the presence of imperfect MR water/fat separation. Specifically, it investigates whether incorporating water/fat information directly or through a maximum likelihood algorithm yields the optimal solution both in terms of reduced crosstalk between oxyhemoglobin and water, and compares results to having no priori knowledge of water and fat.

Fluorescence molecular tomography (FMT) has the potential to become a powerful quantitative research tool for pre-clinical applications such as evaluating the efficacy of experimental drugs. In this paper, we show how a time-domain FMT/microCT instrument can in principle be used to monitor volumetric fluorescence intensity over time for low fluorophore concentration levels. The experimental results we present relate to Protoporphyrin IX which has a quantum efficiency as much as two orders of magnitude lower compared to more conventional extrinsic dyes used for molecular imaging (e.g., Alexa Fluor dyes, Cyanine dyes). Our results highlight the high sensitivity of the single photon counting technology on which the optical system we have built is based. In conjunction with this system we have developed a diffuse optical fluorescence reconstruction technique that is robust and shown here to perform adequately even in cases when the contribution of noise to the data is important. Related to this, we show that the regularization scheme we have developed is reliable even for low fluorophore concentration values and that no adjustment of the regularization parameter needs to be made for different levels of noise. This generic reconstruction approach insures that images reconstructed from data sets acquired at different times and for different fluorescence levels can be compared on an equal footing.

Multi-spectral Near Infrared tomographic imaging has the potential to provide information about patho-physiological function of soft tissue. However, the specific choice of wavelengths used is crucial for the accurate separation of such parameters. It will be demonstrated that the conventionally believed choice of large set of wavelengths can be detrimental in accurate recovery of tissue specific functions. The method of determining a set of optimized bands of wavelengths will be presented and are tested using simulations and experimental data. It will be shown that the optimization method achieves images as accurate as using the full spectrum, but improves crosstalk between parameters. Additionally, a Jacobian normalization technique is presented which takes into account the varying magnitude of different optical parameters within image reconstruction, creating a more uniform update within a spectral image reconstruction model.

The National Institutes of Health (NIH) has recently developed an extensible imaging platform (XIP), a new open-source software development platform. XIP can be used to rapidly develop imaging applications designed to meet the needs of the optical imaging community. XIP is a state-of-the-art set of visual 'drag and drop' programming tools and associated libraries for rapid prototyping and application development. The tools include modules tailored for medical imaging, many of which are GPU hardware accelerated. They also provide a friendlier environment for utilizing popular toolkits such as ITK and VTK, and enable the visualization and processing of optical imaging data and standard DICOM data. XIP has built-in functionality for multidimensional data visualization and processing, and enables the development of independently optimized and re-usable software modules, which can be seamlessly added and interconnected to build advanced applications. XIP applications can run "stand alone", including in client/server mode for remote access. XIP also supports the DICOM WG23 "Application Hosting" standard, which will enable plug-in XIP applications to run on any DICOM host workstation. Such interoperability will enable the optical imaging community to develop and deploy modular applications across all academic/clinical/industry partners with WG23 compliant imaging workstations.

Biophysical models of hemodynamics provide a tool for quantitative multimodal brain imaging by allowing a deeper understanding of the interplay between neural activity and blood oxygenation, volume and flow responses to stimuli. Multicompartment dynamical models that describe the dynamics and interactions of the vascular and metabolic components of evoked hemodynamic responses have been developed in the literature. In this work, multimodal data using near-infrared spectroscopy (NIRS) and diffuse correlation flowmetry (DCF) is used to estimate total baseline hemoglobin concentration (HBT₀) in 7 adult subjects. A validation of the model estimate and investigation of the partial volume effect is done by comparing with time-resolved spectroscopy (TRS) measures of absolute HBT₀. Simultaneous NIRS and DCF measurements during hypercapnia are then performed, but are found to be hardly reproducible. The results raise questions about the feasibility of an all-optical model-based estimation of individual vascular properties.

Although fluorescence imaging has been applied in tumour diagnosis from the early 90s, just the last few years it has met an increasing scientific interest due to the advances in the biophotonics field and the combined technological progress of the acquisition and computational systems. In addition there are expectations that fluorescence imaging will be further developed and applied in deep tumour diagnosis in the years to come. However, this evolving field of imaging sciences has still to encounter important challenges. Among them is the expression of an accurate forward model for the solution of the reconstruction problem. The scope of this work is to introduce a three dimensional coupled radiative transfer and diffusion approximation model, applicable on the fluorescence imaging. Furthermore, the solver incorporates the super-ellipsoid models and sophisticated image processing algorithms to additionally provide a-priori estimation about the fluorophores distribution, information that is very important for the solution of the inverse problem. Simulation experiments have proven that the proposed methodology preserves the accuracy levels of the radiative transfer equation and the time efficacy of the diffusion approximation, while in the same time shows extended success on the registration between acquired and simulated images.

A recent research study has shown that combining multiple parameters, drawn from optical tomographic images, leads to better classification results to identifying human finger joints that are affected or not affected by rheumatic arthritis RA. Building up on the research findings of the previous study, this article presents an advanced computer-aided classification approach for interpreting optical image data to detect RA in finger joints. Additional data are used including, for example, maximum and minimum values of the absorption coefficient as well as their ratios and image variances. Classification performances obtained by the proposed method were evaluated in terms of sensitivity, specificity, Youden index and area under the curve AUC. Results were compared to different benchmarks ("gold standard"): magnet resonance, ultrasound and clinical evaluation. Maximum accuracies (AUC=0.88) were reached when combining minimum/maximum-ratios and image variances and using ultrasound as gold standard.

Prostate cancer diagnosis is based on PSA dosage and digital rectal examination. In case of positive test, a biopsy is conducted and guided by ultrasound imaging. Today, however, as ultrasound imaging is not able to precisely detect tumors, some biopsies have to be performed in the prostate and the only way to improve detection is to increase the number of those uncomfortable biopsies. In order to decrease this number and to improve the patient wellness, we are studying a way to couple ultrasound and fluorescence optical imaging on an endorectal probe. The ultrasounds are used to get morphological information on the prostate and the optical system to detect and to localize fluorophore marked tumors. To support the development of such a system, we have carried out a new tissue-mimicking phantom which represents the three different kind of tissue concerned during prostate endorectal examination: prostate, rectum, surrounding tissues. It was imaged by ultrasound and by fluorescence diffuse optical imaging. We have proved that the optical system is able to detect and to localize a fluorescing inclusion at different depth inside the phantom which has then been superimposed to the morphological image provided by the ultrasounds.

A dual-band trans-rectal optical tomography system is constructed based on an endo-rectal near-infrared/ultrasound applicator that has been developed previously in our laboratory. The endo-rectal NIR/US applicator consists of a commercial bi-plane ultrasound and a NIR probe attached to the sagittal ultrasound transducer. The NIR probe consists of 7 illumination & 7 detection channels that are distributed in parallel to and aside the sagittal TRUS transducer. The emissions from a 780nm and an 830nm light sources are combined and delivered sequentially to the 7 NIR source channels of the endo-rectal NIR/US probe. The 7 NIR detection channels are coupled to a spectrometer for separation of the signals at two wavelengths illuminated from single source channel. The dual-band signals from all source channels are acquired sequentially by a CCD camera synchronized with the source switching. The acquisition of dual-band trans-rectal optical tomography data is accompanied by position-correlated concurrent trans-rectal ultrasound imaging. The reconstruction of a target at dual-wavelength illumination is guided by a priori spatial information provided by the sagittal trans-rectal ultrasound. Liquid phantoms with different hemoglobin concentration and oxygen saturation are used to test the feasibility of dual-band trans-rectal optical tomography.

Tissue phantoms simulating the human breast were used to demonstrate the imaging capabilities of an MRI-coupled fluorescence molecular tomography (FMT) imaging system. Specifically, phantoms with low tumor-to-normal drug contrast and complex internal structure were imaged with the MR-coupled FMT system. Images of indocyanine green (ICG) fluorescence yield were recovered using a diffusion model-based approach capable of estimating the distribution of fluorescence activity in a tissue volume from tissue-boundary measurements of transmitted light. Tissue structural information, which can be determined from standard T1 and T2 MR images, was used to guide the recovery of fluorescence activity. The study revealed that this spatial guidance is critical for recovering images of fluorescence yield in tissue with low tumor-to-normal drug contrast.

MicroRNAs (miRNAs) are one of the most prevalent small (~22 nucleotide) regulatory RNA classes in animals. These miRNAs constitute nearly one percent of genes in the human genome, making miRNA genes one of the more abundant types of regulatory molecules. MiRNAs have been shown to play important roles in cell development, apoptosis, and other fundamental biological processes. MiRNAs exert their influence through complementary base-pairing with specific target mRNAs, leading to degradation or translational repression of the targeted mRNA. We have identified and tested a novel microRNA (miR-491) and demonstrated increased apoptosis in hepatocellular carcinoma cells (HepG2) and in human breast cancer cells (HBT3477) in vitro. We prepared a novel cancer targeting assembly of gold nanoparticles (GNP) with Quantum dots, miR-491, and MAb-ChL6 coupled through streptavidin/biotin for effective transfection, and to induce apoptosis in specific cancer cells for imaging and targeted therapy. The targeting and apoptosis inducing ability was tested by confocal and electron microscopy. The MAb-GNP-miR491-Qdot construct effectively transfected into the HBT3477 cells and induced apoptosis the confirmation of these results would suggest a new class of molecules for the imaging and therapy of breast cancer.

Imaging the structure and correlating it with the biochemical content of the retina holds promise for fundamental research and for clinical applications. Optical coherence tomography (OCT) is commonly used to image the 3D structure of the retina and while the added functionality of biochemical analysis afforded by Raman scattering could provide critical molecular signatures for clinicians and researchers, there are many technical challenges to combining these imaging modalities. We present an ex vivo OCT microscope combined with Raman spectroscopy capable of collecting morphological and molecular information about a sample simultaneously. The combined instrument will be used to investigate remaining technical challenges to combine these imaging modalities, such as the laser power levels needed to achieve a Raman signal above the noise level without damaging the sample.

Blood oxygenation level dependent (BOLD) response, which is measured by functional magnetic resonance imaging (fMRI), is known to be a combination of various vascular parameters, among which deoxy-hemoglobin is argued to be a major contributor. Functional near infrared spectroscopy (fNIRS), though being limited in its spatial resolution, provides a promising tool to study cortical activations, due to its specificity of independent measurement of blood parameters (Oxy, De-oxy and Total Hemoglobin), high temporal resolution and ease of use. To study the close relationship between these imaging modalities, a finger tapping task with stimulus durations (2, 4, 8 & 16 sec) with variable inter-stimulation intervals was chosen to compare spatio-temporal properties and non-linearity of BOLD signal with HbO, HbR and HBT signal. This helped determine what parameter (HbO, HbR and HbT) BOLD signals correlate to most and how factors like neural adaptation that cause non-linearity can affect the hemodynamic behavior. It investigates the non-linearity in HbO, HbR and HbT concentrations as compared to BOLD signal obtained using simultaneous fNIRS and fMRI measurement. Investigating non-linearity in hemodynamic response could provide a better understanding of neuronal function by modeling neural adaptation. The paper also discusses a method to model the neural adaptation and hemodynamic response.

We present an instrument for simultaneous imaging of the rodent brain with frequency-domain optical tomography and magnetic resonance imaging. The instrument uses a custom-built fiber optic probe that allows for measurements in backreflectance geometry. The probe consists of 13 source and 26 detector fibers and is small enough to fit inside of a microMRI RF coil with an inner diameter of 38mm. Illumination by the source fibers is time demultiplexed by an optical fiber switch. A gain-modulated image intensifier CCD camera focuses onto the endpoints of large-core gradedindex detector fibers and collects the frequency-domain data. Imaging can be performed with source-modulation frequencies up to 1 GHz. The instrument is capable of acquiring multi-frequency optical tomography data at 2 wavelengths, and the data can be used to generate 3D maps of hemoglobin concentrations. At the same time magnetic resonance images can be acquired with in-plane resolution smaller than 100 micron. To illustrate the performance of the instrument we show results of small animal studies that involve inhalation of 100% carbogen and chemically induced seizures.

An approach of hierarchically implementing the spatial prior information in trans-rectal optical tomography is introduced. Trans-rectal optical imaging of the prostate deals with photon propagation through the rectum wall, the peri-prostate tissue and the prostate. Reconstructing a lesion in the prostate is challenging due to the structural complexity as well as the optical heterogeneity. Incorporating spatial "hard" a priori information available from complementary imaging modalities such as trans-rectal ultrasound could in principle improve the accuracy of trans-rectal optical tomography reconstruction. However, the reconstruction is potentially subject to the local-minimum sensitivity if the values of all regional optical properties are to be initialized simultaneously. We propose a hierarchical spatial prior approach for trans-rectal optical tomography reconstruction. Instead of assigning the initial values to all sub-regions at once, a region is initially assumed homogenous, and the reconstructed optical properties are used as the initial guess for the region as a background when a sub-region is included in the next step. This approach translates to a 3-step iteration routine whereby the first step reconstructs the entire imaging volume as a single region, the second step uses these results as the initial guess of peri-prostate tissue to reconstruct the prostate and the rectum wall, and the third step assigns the updated results as the initial values of 3 existing regions to reconstruct a lesion inside the prostate. This approach, validated by simulation and applied to experimental measurements, is more reliable in global convergence, robust in imaging of single or multiple targets, and accurate for the recovering of optical properties.

We report an all-fiber system aided by double-clad fiber (DCF) and DCF devices for simultaneous measurements of optical coherence tomography (OCT) and fluorescence spectroscopy (FS). The DCF together with DCF coupler and single-body DCF lens helped in realizing a multifunctional single-unit probe for the OCT-FS system. The fiber lens formed on the DCF aids in effective focusing and signal collection, while the DCF coupler collects the OCT signal from the core and the fluorescence signal from the cladding of the DCF. The OCT image and fluorescence spectra of plant tissues are simultaneously measured and presented to validate the performance.

The temperature of human body on the surface of the skin depends on the metabolic activity, the blood flow, and the temperature of the surroundings. Any abnormality in the tissue, such as the presence of a tumor, alters the normal temperature on the skin surface due to increased metabolic activity of the tumor. Therefore, abnormal skin temperature profiles are an indication of diseases such as tumor or cancer. This study is to present an approach to detect the female breast tumor and its related parameter estimations by combination the finite element method with infrared thermography for the surface temperature profile. A 2D simplified breast embedded a tumor model based on the female breast anatomical structure and physiological characteristics was first established, and then finite element method was used to analyze the heat diffuse equation for the surface temperature profiles of the breast. The genetic optimization algorithm was used to estimate the tumor parameters such as depth, size and blood perfusion by minimizing a fitness function involving the temperature profiles simulated data by finite element method to the experimental data obtained by infrared thermography. This preliminary study shows it is possible to determine the depth and the heat generation rate of the breast tumor by using infrared thermography and the optimization analysis, which may play an important role in the female breast healthcare and diseases evaluation or early detection. In order to develop the proposed methodology to be used in clinical, more accurate anatomy 3D breast geometry should be considered in further investigations.

In this paper a novel method for determining refractive indices of a multi-layered samples using low coherence interferometry (LCI), developed at the National Physical Laboratory, UK, is introduced. Conventional Optical Coherence Tomography (OCT) utilises a lateral scanning optical probe beam to construct a depth resolved image of the sample under investigation. All interfaces are detected in optical path length, resulting in an image depending on the refractive index of all prior layers. This inherent ambiguity in optical and geometric path length reduces OCT images to purely qualitative ones. We have demonstrated that by optically probing the sample at multiple angles we can determine bulk refractive index of layers throughout plane parallel samples. This method improves upon current approaches of extracting refractive index parameters from multi-layered samples as no prior geometrical information is required of the sample and the phase index for each layer is obtained as opposed to the group index. Consequently the refractive index result for each layer is independent of the refractive index of surrounding layers. This technique also improves on conventional measurements, as it is less susceptible to error due to surface defects. This technique is easily implemented, and can easily be modified to obtain in situ measurements. Investigating a silica test piece and comparing the refractive index obtained by that of standard critical angle refractometry has validated the robustness of the technique.

Heat is one of the most important parameters of living beings. Skin temperature is not the same on the entire body and so, a thermal signature can be got. Infrared map on serial imaging can constitute an early sign of an abnormality. Thermography detects changes in tissue that appear before and accompany many diseases including cancer. As this map has a better resolution an early cancer diagnosis can be done. The temperature of neoplasic tissue is different up to 1.5 °C than that of the healthy tissue as a result of the specific metabolic rate. The infrared camera images show very quickly the heat transferred by radiation. A lot of factors disturb the temperature conversion to pixel intensity. A sensitive temperature sensor with a 10 Mpixels video camera, showing its spatial position, and a computer fusion program were used for the map with high spatial-temperature resolution. A couple of minutes are necessary to get a high resolution map. The asymmetry and borders were the main parameters analyzed. The right cancer diagnosis was for about 78.4% of patients with thyroid cancer, and more than 89.6% from patients with breast cancer. In the near future, the medical prognosis will be improved by fractal analysis.

Retinal nerve fiber layer (RNFL) thickness, a measure of glaucoma progression, can be measured in images acquired by spectral domain optical coherence tomography (OCT). The accuracy of RNFL thickness estimation, however, is affected by the quality of the OCT images. In this paper, a new parameter, signal deviation (SD), which is based on the standard deviation of the intensities in OCT images, is introduced for objective assessment of OCT image quality. Two other objective assessment parameters, signal to noise ratio (SNR) and signal strength (SS), are also calculated for each OCT image. The results of the objective assessment are compared with subjective assessment. In the subjective assessment, one OCT expert graded the image quality according to a three-level scale (good, fair, and poor). The OCT B-scan images of the retina from six subjects are evaluated by both objective and subjective assessment. From the comparison, we demonstrate that the objective assessment successfully differentiates between the acceptable quality images (good and fair images) and poor quality OCT images as graded by OCT experts. We evaluate the performance of the objective assessment under different quality assessment parameters and demonstrate that SD is the best at distinguishing between fair and good quality images. The accuracy of RNFL thickness estimation is improved significantly after poor quality OCT images are rejected by automated objective assessment using the SD, SNR, and SS.

In cancer diagnostics the most important problems are the early identification and estimation of the tumor growth and spread in order to determine the area to be operated. The aim of the work was to design of statistical algorithms helping doctors to objectively estimate pathologically changed areas and to assess the disease advancement. In the research, algorithms for classifying endoscopic autofluorescence images of larynx and intestine were used. The results show that the statistical pattern recognition offers new possibilities for endoscopic diagnostics and can be of a tremendous help in assessing the area of the pathological changes.

Visualization of the internal structures of the retina is critical for clinical diagnosis and monitoring of pathology as well as for medical research investigating the root causes of retinal degeneration. Optical Coherence Tomography (OCT) is emerging as the preferred technique for non-contact sub-surface depth-resolved imaging of the retina. The high resolution cross sectional images acquired in vivo by OCT can be compared to histology to visually delineate the retinal layers. The recent demonstration of the significant sensitivity increase obtained through use of Fourier domain (FD) detection with OCT has been used to facilitate high speed scanning for volumetric reconstruction of the retina in software. The images acquired by OCT are purely structural, relying on refractive index differences in the tissue for contrast, and do not provide information on the molecular content of the sample. We have constructed a FDOCT prototype and combined it with a fluorescent Scanning Laser Ophthalmoscope (fSLO) to permit real time alignment of the field of view on the retina. The alignment of the FDOCT system to the specimen is crucial for the registration of measurements taken throughout longitudinal studies. In addition, fluorescence detection has been integrated with the SLO to enable the en face localization of a molecular contrast signal, which is important for retinal angiography, and also for detection of autofluorescence associated with some forms of retinal degeneration, for example autofluorescence lipofuscin accumulations are associated with Stargardt's Macular Dystrophy. The integrated FD OCT/fSLO system was investigated for imaging the retina of the mice in vivo.

Multispectral imaging is extensively used to multiplex the labeling of tissue features for diagnostic purposes. This technique enables multiple biomarkers with overlapping spectra to be distinguished using unmixing algorithms. However, spectral information alone is sometimes not sufficient enough to obtain perfect fluorochrome separation. By collecting time-resolved multispectral data containing a fluorescence decay time-series at each wavelength, the added temporal information can be used along with the spectral measurements as an additional distinguishing factor between the different dyes. A monoexponential decay model is used to segment the image into pure and mixed pixels based on the fit residuals. Using this information, the lifetimes and spectra of the dyes are retrieved using a global parameter estimation model within the segmented regions. The image is unmixed using the retrieved spectra and lifetimes of the individual components.

Diffuse optical tomography using perturbation Monte Carlo method can overcome the drawback inherent to the classical model based on the diffusion equation for pre-clinical applications. The combined use of information from different time gates enables image reconstruction with accurate quantification and reducing the crosstalk between the absorption and the scattering coefficients. In this work, we apply this approach to solve an inverse problem of a 3D mouse model, and investigate the benefit to incorporate the time-resolved data into the optical reconstruction for quantitative functional imaging.

We propose the use of Time-Resolved Diffuse Optical Tomography in a multispectral scheme with anatomical constraints supplied by MR imaging to reconstruct functional parameters of the animal model with greater accuracy and resolution. The tomographic imaging system described is capable of acquiring temporal measurements in multiple-views using a gated ICCD camera. A tunable Ti-Sapphire pulsed laser at wavelengths between 700nm - 1000nm is used as the source. Anatomical distribution is determined using MRI in a non-concurrent setting. Time-resolved measurements at multiple wavelengths in the NIR window combined with the anatomical constraints is used to determine a 3D distribution of the functional parameters in vivo. Multispectral spectroscopy measurements on homogenous tissue simulating phantoms are used to demonstrate the accuracy of the system in determining optical parameters in thin tissues. We show that temporal measurements combined with MRI data can be used to accurately quantify optical properties in heterogeneous tissues.