This PDF file contains the front matter associated with SPIE Proceedings Volume 7174, including the Title Page, Copyright information, Table of Contents, and the Conference Committee listing.

We show the performance of a proposed perturbation theory of the diffusion equation for studying light propagation in optically heterogeneous media, i.e. characterized by a distribution of the absorption and the reduced scattering coefficient. In different geometries (cylindrical, slab, layered), we study the change of continuous wave intensity due to the presence of focal absorption perturbations. The results obtained with fourth order perturbation theory show a clear improved accuracy with respect to first order calculations for a range of the absorption contrasts of interest in the field of near infrared spectroscopy and diffuse optical tomography. The method of Padè Approximants is used to extend the limits of the proposed perturbation theory to a wider range of absorption contrasts. For validation of the theory, we show comparison with Monte Carlo simulations.

The equation of radiative transfer (ERT) is generally accepted as the most accurate model for light propagation in biological tissues. The ERT is notoriously expensive to solve numerically. Recently, Klose and Larsen have approximated the time-independent ERT using the simplified spherical harmonics equations ( SP_N approximation). In this work, we outline how to derive the SP_N approximation of the time-dependent ERT and obtain the associated integro- partial differential equations involving temporal convolution integrals. No approximation is made as regards the time variable in our derivation. To simplify the numerical solution of these equations, we introduce a "memory function". We discuss the numerical solution for N = 1 in the 2D and homogeneous case. We provide time evolution maps of the solution and compare it with the diffusion approximation of the ERT. The findings presented here straightforwardly extend to 3D inhomogeneous media and for higher values of N. These more complicated cases along with further details will be reported elsewhere.

The geometry of trans-lumenal diffuse optical measurement is considerably different from that of externally applied diffuse optical imaging. In externally-applied diffuse optical imaging of breast, brain, etc, an analytic solution to the diffusion equation for a planar semi-infinite medium is often applied. This solution works accurately for planar applicator and is a good approximation for a ring applicator of considerable size. In trans-lumenal diffuse optical imaging of internal organs like the prostate, the applicator likely should have a convex surface profile for interfacing with a typically circular cross-section of the lumen. The influence of this convex applicator shape upon the photon transport is expected to cause a deviation from the solution predicted by a semi-infinite planar boundary. This interference, if available, is particularly relevant to the axial geometry in trans-lumenal diffuse optical imaging. This work investigates the analytic solution of continuous-wave photon diffusion equation for axial imaging when a cylindrical trans-lumenal applicator is used. The Green's function of the photon diffusion equation in an infinite medium geometry is expanded in cylindrical coordinates, and an image-source method is utilized to derive the analytic solution for circular concave & circular convex boundary profiles based on extrapolated boundary condition. Numerical evaluations are conducted to examine the effect of the circular boundary. Empirical solution potentially useful for calibrating the photon remission data in a circular boundary is also derived. The numerical evaluation results and the empirical solution are subject to validation against Monte Carlo simulations and experimental measurements.

Radiative transport in the delta-P1 and delta-P3 approximations with an extrapolated boundary condition combined with the reciprocity was demonstrated providing accurate estimations for optical perturbations by comparing with experimental data in a small volume of a liquid tissue phantom. For an absorbing target submerged in a homogeneous medium with an albedo of 0.75, both approximations estimated the optical perturbation with acceptable accuracy when the depth of the target is larger than 1.5 mm. For a fluorescent target, the two approximations estimated the perturbations with high accuracy for both shallow and deep regions. Therefore, these two approximations combined with the reciprocity were proposed for rapid tomographic image reconstruction in a small volume or superficial tissues.

Computational speed and available memory size on a single processor are two limiting factors when using the frequency-domain equation of radiative transport (FD-ERT) as a forward and inverse model to reconstruct three-dimensional (3D) tomographic images. In this work, we report on a parallel, multiprocessor reducedspace sequential quadratic programming (RSQP) approach to improve computational speed and reduce memory requirement. To evaluate and quantify the performance of the code, we performed simulation studies employing a 3D numerical mouse model. Furthermore, we tested the algorithm with experimental data obtained from tumor bearing mice.

In this work, we explore diffuse correlation spectroscopy (DCS) in a two-layered geometry. We compare the effiency of an homogeneous and a two-layered model to recover flow changes. By simulating a realistic human head with MRI anatomical data, we show that the two-layered model allows distinction between superficial layers and brain hemodynamic changes. The results show that the two-layered model provides a better estimate for the flow change than the homogeneous one. Experimental measurements with a two-layered dynamical phantom confirm the ability of the two-layered analytical model to distinguish flow increase in each layer.

This study describes noninvasive noncontact methods to acquire and analyze functional information from the skin. Multispectral images at several selected wavelengths in the visible and near infrared region are collected and used in mathematical methods to calculate concentrations of different chromophores in the epidermis and dermis of the skin. This is based on the continuous wave Near Infrared Spectroscopy method, which is a well known non-invasive technique for measuring oxygenation changes in the brain and in muscle tissue. Concentration changes of hemoglobin (dO₂Hb, dHHb and dtHb) can be calculated from light attenuations using the modified Lambert Beer equation. We applied this technique on multi-spectral images taken from the skin surface using different algorithms for calculating changes in O₂Hb, HHb and tHb. In clinical settings, the imaging of local oxygenation variations and/or blood perfusion in the skin can be useful for e.g. detection of skin cancer, detection of early inflammation, checking the level of peripheral nerve block anesthesia, study of wound healing and tissue viability by skin flap transplantations. Images from the skin are obtained with a multi-spectral imaging system consisting of a 12-bit CCD camera in combination with a Liquid Crystal Tunable Filter. The skin is illuminated with either a broad band light source or a tunable multi wavelength LED light source. A polarization filter is used to block the direct reflected light. The collected multi-spectral imaging data are images of the skin surface radiance; each pixel contains either the full spectrum (420 - 730 nm) or a set of selected wavelengths. These images were converted to reflectance spectra. The algorithms were validated during skin oxygen saturation changes induced by temporary arm clamping and applied to some clinical examples. The initial results with the multi-spectral skin imaging system show good results for detecting dynamic changes in oxygen concentration. However, the optimal algorithm needs to be determined. Multi-spectral skin imaging shows to be a promising technique for various clinical applications were the local distribution of oxygenation is of major importance.

Near-infrared spectroscopy (NIRS) calculates hemoglobin parameters, such as oxygenated hemoglobin (oxyHb) and deoxygenated hemoglobin (deoxyHb) using the near-infrared light around the wavelength of 800nm. This is based on the modified-Lambert-Beer's law that changes in absorbance are proportional to changes in hemoglobin parameters. Many conventional measurement methods uses only a few wavelengths, however, in this research, basic examination of NIRS measurement was approached by acquiring wide range of wavelength information. Venous occlusion test was performed by using the blood pressure cuff around the upper arm. Pressure of 100mmHg was then applied for about 3 minutes. During the venous occlusion, the spectrum of the lower arm muscles was measured every 15 seconds, within the range of 600 to 1100nm. It was found that other wavelength bands hold information correlating to this venous occlusion task. Technique of improving the performance of NIRS measurement using the Spectroscopic Method is very important for Brain science.

Optical topography is a brain function imaging technology based on near-infrared spectroscopy, whereby changes in oxygenated hemoglobin (oxy-Hb) and deoxygenated hemoglobin (deoxy-Hb) are observed. There are several components other than hemoglobin in biomedical tissue such as cytochrome oxidase (cyt-ox) and water, and it is important to measure them. We investigated the possibilities of multi-wavelength measurement of the changes in oxy- and deoxy-Hb, the redox state of cyt-ox, and water based on a modified Beer-Lambert law. According to error propagation theory, we searched for the wavelength combination where the estimation errors were minimal in the wavelength range from 650 to 900 nm under several conditions of measured components and number of wavelengths. Next, using the experimental data, we performed the same search. The signal-to-noise ratio (SNR) loss depended on measured components and wavelengths used in the measurement. We calculated the necessary improvement in precision of absorbance measurement when we estimated the changes in the redox state of cyt-ox and water, for maintaining the same level of SNRs in the hemoglobin-only measurement. The measurements of both the redox state of cyt-ox and water were not practical because the SNR losses of Hb became more than 30 dB, while the measurement where only one of them was performed was more feasible. We calculated the required ratio of absorbance change to its noise for the acquisition of cyt-ox at the same SNR of oxy-hemoglobin in hemoglobin-only measurements when the absorbance change caused by cyt-ox was assumed to be one tenth of that of oxy-hemoglobin.

The clinical motivation for our work was to help surgeons see vessels through non-translucent intraoperative tissues during laparoscopic removal of the gallbladder. Our main focus was to answer the question of how CW imaging performs relative to ICCD (Intensified Charge-Coupled Device) based time-gated imaging, which is a lot more costly, under broad Gaussian beam illumination conditions. We have studied the simplified case of an isolated bile duct embedded at different depths within a 2 cm slab of adipose tissue. Monte Carlo simulations were preformed for both reflectance and trans-illumination geometries. The relative performance of CW versus time-gated imaging was compared in terms of spatial resolution and vessel detection sensitivity in the resulting simulated images. Experiments were performed in reflectance geometry to validate simulation results. It was found that time-gated imaging offers superior spatial resolution and vessel detection sensitivity in all cases though CW trans-illumination measurements may also offer satisfactory performance for this tissue geometry at a lower cost.

Most high-resolution optical tomography techniques employ coherence domain or time domain methodologies to capture non-scattered photons in turbid media. Angular Domain Optical Projection Tomography (ADOPT) uses an angular filter array (AFA) to observe photons that propagate through a specimen with small angular deviation. We constructed an ADOPT system consisting of an AFA micro-machined silicon micro-tunnel array with each micro-tunnel 60 μm wide, 60 μm high, 10 mm long, and separated by 5 μm thick walls. The range of acceptance angles was 0° to 0.5°. The system also included an 808 nm CW diode laser, beam shaping optics, a sample cuvette, a Keplerian lens system, and a CMOS camera. Testing was performed with a target consisting of two graphite rods (0.9 mm diameter) suspended in the cuvette by a rotation stage. The target was placed in a manner that the line of laser light was perpendicular to the long axis of the rods. A multitude of projections were collected at increments of 1.8° and compiled into a sinogram. A transverse image was reconstructed from the sinogram using filtered backprojection. The submillimeter targets embedded in the 2 cm thick scattering medium (reduced scattering coefficient ≤ 2.4 cm^-1) were discernable in both the sinograms and the reconstructed images. The results suggest that ADOPT may be a useful technique for tomographic imaging of thick biological specimens (i.e. up to 8 mm across).

The use of spatially modulated light is finding application in biomedical optics having potential use in imaging and tomography of tissues and small animals. We describe the time-resolved propagation of spatial frequencies in turbid media. We present a set-up based on a ps laser source, spatially modulated by a micro-mirror device and a time-gated intensifier. We discuss the relevant information content that can be useful for imaging of tissues, in terms of the spatial Fourier components of the propagating pulse. We demonstrate that high spatial frequencies appear in the early time-gated signal whereas low frequencies persist for longer times and that the combined use of high spatial frequencies and early time gates can be used to improve the resolution in imaging.

Thus far optical imaging of the finger joints is limited to only one joint at a time. We have developed a CCD-based system that can image a whole finger, covering multiple joints. We show that the two joints of a volunteer's index finger can be imaged clearly and concurrently. The comparison of the optical images with the x-ray image of the same finger indicates that the recovered size of the whole finger, the bones and the two joint spaces are consistent with the findings from the x-ray. This study demonstrates a practicable system for full finger imaging and indicates the possibility of full hand imaging if a fast laser source scanning subsystem is integrated with the system.

It is well known that diffuse optical tomography (DOT) has limited spatial resolution, and sufficient contrast recovery is limited to lesions greater than ~6 mm[1]. However, with the addition of multimodality methods that combine high spatial resolution imaging, such as MRI, it has been shown that quantification and feature recovery improves[2]. However, it is not known how well MRg-DOS will perform with characterizing small lesions in 3D. These limits need to be established in order to determine the practical limitations of optical imaging. This paper investigates the contrast resolution limits of 3 dimensional MRg-DOS. Short irregular inclusions of various diameters are added to a homogeneous background. Two case studies are presented which represent these limiting situations.

We present a wavelength optimization approach incorporating the spectral derivative method in near infrared (NIR) reconstruction. The similarity between spectral features of oxy-hemoglobin and water leads to difficulty in separating them with any spectral reconstruction method. By taking the difference of wavelength pairs and using a spectral derivative fitting method, this similarity can be broken down to better differentiate each set of two chromophores. Optimal wavelength permutations were chosen based on the criteria of minimum correlation of the differences in adjacent wavelengths. The direct and derivative methods were then compared with optimal wavelength sets in two and three dimensional reconstructions. The wavelength optimized derivative method shows superior results in the recovery of chromophore concentrations. This approach can not only yield reduced coupling and geometry errors, it also shows better separation of one chromophore from others.

We have combined hyperspectral imaging with spectral domain optical coherence tomography (SDOCT) to non-invasively image changes in hemoglobin saturation, blood flow, microvessel morphology and sheer rate on the vessel wall with tumor growth. Changes in these hemodynamic variables were measured over 24 hours in dorsal skin fold window chamber tumors. There was a strong correlation between volumetric flow and hemoglobin saturation (ρ = 0.89, p = 9 × 10^-6, N = 15), and a moderate correlation between shear rate on the vessel wall and hemoglobin saturation (ρ = 0.56, p = 0.03, N=15).

A multi-modality imaging approach and instrument that integrate optical imaging system and near-infrared spectroscopy into an x-ray tomosynthesis setup have been employed to perform a clinical study of image-guided spectroscopy on osteoarthritis (OA) in the finger joints. The multiwavelength spectroscopy of the joints using x-ray-guided spatial constraints provides 3D images of oxygen saturation and water content with high resolution and improved quantitative capability. Based on the recovered quantitative results from 18 healthy volunteers and 22 patients, we observed that oxygen saturation and water content were significant discriminators for differentiation of healthy joints from diseased ones. The recovered images appear to show that the OA joints have high water values and decreased oxygen saturation.

We present an experimental study on four rabbits to demonstrate the feasibility of near-infrared spectroscopy in the noninvasive assessment of testicular torsion. We used a multi-distance frequency-domain method, based on a fixed detector position and a 9-mm linear scan of the illumination optical fibers, to measure absolute values of pre- and post-operative testicular oxygen saturation. Unilateral testicular torsions (by 0°, 540° or 720°) on experimental testes and contralateral sham surgeries (no torsion) on control testes were performed and studied. Our results showed (a) a consistent baseline absolute tissue oxygen saturation value of 78% ± 5%; (b) a comparable absolute saturation of 77% ± 6% on the control side (testes after sham surgery); and (c) a significantly lower tissue oxygen saturation of 36% ± 2% on the experimental side (testes after 540° or 720° torsion surgery). These results demonstrate the capability of frequency domain nearinfrared spectroscopy in the assessment of absolute testicular hemoglobin desaturation caused by torsion, and show promise as a potential method to serve as a complement to conventional color and spectral Doppler ultrasonography.

We developed a hybrid continuous-wave/frequency-domain instrument to obtain both spatial and spectral information of the female breast. The two-dimensional (2D) tandem planar scanning of a compressed breast enables a pixel size of 2×2 mm² and a continuous spectra acquisition from 650 nm to 900 nm at every image pixel with a 0.5 nm spectral step. A 2D spline interpolation algorithm is implemented to increase the data sampling rate and reduce the pixel size to 0.5×0.5 mm². We then apply an edge-correction method to compensate the signal change due to the breast thickness variation. The resulted optical density image is further processed using a previously developed second-derivative algorithm to enhance the contrast and improve the spatial resolution of the optical inhomogeneities within the breast tissue. The finer structures displayed in the second-derivative image offer better identification of the pixels of interest associated with significant hemoglobin presence. We then employ a novel paired-wavelength oximetry method to determine the absolution value of oxygen saturation for those identified pixels of interest. We found the majority of oxygen saturation values from two healthy human subjects fall within the range 60%-95%, which is consistent with previously published results. Breast oximetry could have a potential applicability toward breast cancer detection and diagnostics and this novel paired-wavelength method can be a robust and accurate way to retrieve the oxygenation information in vivo.

Breast cancer characteristics such as angiogenesis and hypoxia can be quantified by using optical tomography imaging to observe the hemodynamic response to an external stimulus. A digital near-infrared tomography system has been developed specifically for the purpose of dynamic breast imaging. It simultaneously acquires four frequency encoded wavelengths of light at 765, 808, 827, and 905nm in order to facilitate the functional imaging of oxy- and deoxy-hemoglobin, lipid concentration and water content. The system uses 32 source fibers to simultaneously illuminate both breasts. There are 128 detector fibers, 64 fibers for each breast, which deliver the detected light to silicon photo-detectors. The signal is conditioned by variable gain amplifiers and filters and is quantized by an analog to digital converter (ADC). The sampled signal is then passed on for processing using a Digital Signal Processor (DSP) prior to display on a host computer. The system can acquire 2.23 frames per second with a dynamic range of 236 dB.

An NIRS tomography system that can simultaneously acquire data at three wavelengths has been developed to measure changes in physiological properties with 15 second time resolution. The results of homogenous and heterogeneous blood phantom studies indicated that the R² values between average estimated total hemoglobin (Hb_T) values and blood concentrations are 0.99 and 0.9, respectively. In preliminary normal subject clinical trials, a cohort of normal subjects were tested by acquiring the series images as well as the pressure is adding to and releasing from the breast. The recovered data shown that by adding measurable pressure, Hb_T is reduced and the maximum HbT reduction is correlated to the Body Mass Index.

The chest-wall underneath the breast tissue distorts the diffused near infra-red light measured at distant source-detector pairs. Common image reconstruction method consider the media as homogeneous and applying the semi-infinite model. In this paper, we have compared the performance of our two-layer model with semi-infinite model by simulation and a clinical case. The results show that when the chest wall has significant effect on the measurement data, a benign lesion with low absorption can be misled as a malignant case with high absorption by using semi-infinite model. We have also shown the influence of mismatch geometry of breast tissue and chest-wall at lesion and reference sides on the reconstructed image and a correction method has been introduced to reduce these effects. With the assistance of two orthogonal co-registered ultrasounds, the geometry of the breast tissue and chest wall interface can be determined and modeled as a two-layer medium with 3D finite element mesh. Since numerical algorithms based on finite element methods (FEM) are suitable for complex geometry and boundary conditions, this method is adapted to model the chestwall. Four parameters of bulk absorption and reduced scattering coefficients of the first and second layers are estimated and used to characterize the optical properties of the medium. We used a finite element model based on modified born approximation for image reconstruction. A mismatch correction algorithm has been applied to compensate the mismatch geometry of the breast tissue and chest-wall interface at the reference and the lesion side.

We present a method to enhance tumor detectability in breasts imaged with our optical fluorescence mammography system. During a measurement, transmission data at 4 wavelengths and fluorescence data for excitation at 1 wavelength are collected after injection of an optical contrast agent. The data are reconstructed into 3D images of the absorption and fluorescence distributions. Combining those images enables the identification of various breast tissue compounds. Here, we investigate the relevance of our method in phantom experiments.

We report our continued study of phase-contrast diffuse optical tomography (PCCDOT) for evaluating its fidelity in distinguishing malignant breast lesions form benign ones. 144 breast masses were examined from 134 patients, aging from 22~82 with the mean age of 56. Tissue optical parameters including refractive index, and absorption and scattering coefficients, were obtained and compared with their corresponding biopsy/pathology reports. In consistent with our previous study, malignant masses tended to have a decreased refractive index relative to their surrounding normal tissue, which acts as the key character to differentiate them from benign masses. The results show that the specificity is improved significantly over the previous smaller scale study (85% vs. 70%) due to the addition of significantly more benign cases, while the sensitivity stays about the same (81% vs. 82%) due to the similar number of malignant cases used compared to the smaller scale study.

Fluorescent contrast agents have been proposed for near-infrared (NIR) diagnostic breast imaging problems as the most efficient process for inducing optical contrast. We have developed a penalty modified barrier function method (PMBF) with constrained truncated Newton and trust region methods (CONTN) for fluorescence-enhanced NIR diagnostic imaging for both noncontact (area illumination/collection) and contact (point illumination/collection) measurement techniques. A simple logarithmic penalty function has been used in the PMBF/CONTN algorithm with linear convergence. The motivation of this paper is to compare the efficiency and performance of this method with many alternative penalty barrier function methods, (a) hyperbolic penalty function, (b) quadratic-logarithm penalty function, (c) logarithmic exponential penalty function for proper selection of a penalty function that will be the most suitable for optical reconstruction problems. In this paper a numerical comparison of different penalty function methods has been made using experimental measured data in a clinically relevance volume in three-dimensions. Our objective is to continue development of sophisticated constrained optimization PMBF/CONTN method to provide high resolution three dimensional tomography a reality.

We developed a method for solving the fluorescence equation of radiative transfer in the frequency domain on blockstructured grids. In this way fluorescence light propagation in arbitrarily shaped tissue can be modeled with high accuracy without compromising on the convergence speed of these codes. The block-structure grid generator is developed as a multi-purpose tool that can be used with many numerical schemes. We present results from numerical studies that show that it is possible to resolve curved boundaries with grids that maintain much of the intrinsic structure of Cartesian grids. The natural ordering of this grid allows for simplified algorithms. In simulation studies we found that we can reduce the error in boundary fluence by a factor of five by using a two-level block structured grid. The increase in computational cost is only two-fold. We compare benchmark solutions to results with various levels of refinement, boundary conditions, and different geometries.

Depth-resolved imaging of fluorescent molecules in tissue using a non-invasive optical modality called fluorescent molecular tomography (FMT) has found applications in pre-clinical and clinical studies. While FMT offers unique and affordable functional imaging capabilities, its resolution is limited due to the diffusive nature of light propagation in tissue. In this paper we offer a framework for investigating the resolution of FMT using information-theoretic concepts. Specifically, we analyze the amount of useful information that exists in a set of emission measurements. The information content of the measurements directly affects the actual resolution that can be achieved in the reconstructed threedimensional fluorescence images. The relationship between this information content and the measurement geometry is further discussed where it is shown that expanding the measurement size does not necessarily increase the information content. The concept of capacity as defined for multi-input multi-output channels is applied to the linear model of FMT. Assuming a uniform non-zero a priori probability distribution for the fluorophore concentrations in the volume voxels, we derive an expression for the information capacity of the FMT system matrix. This capacity essentially indicates an upper limit on the amount of data that can be extracted from emission measurements. The capabilities of various detector configurations in resolving fluorescent tubes inserted in a gel-based tissue phantom are analyzed in a continuous-wave FMT system using the proposed framework. It is observed that the information capacity of source-detector configurations of different scales directly affects the performance in terms of resolution in the reconstructed fluorescent images.

We have designed a three dimensional (3D) fluorescence optical tomography system for small animal imaging based on an innovative system geometry that uses a truncated conical mirror which permits the entire surface of the animal to be viewed simultaneously by a single CCD camera. Compared with traditional approaches that employ a flat mirror, the conical mirror system has approximately 3 times better measurement sensitivity. By utilizing a fast switching filter wheel (switching time < 100 milliseconds), emission data at multiple wavelengths can be efficiently collected. An array of appropriately shaped neutral density filters, mounted on a linear stage, can be used to increase the system measurement dynamic range by 3 orders of magnitude. An x-y galvo mirror scanning system makes it possible to scan a collimated laser beam to any location on the mouse surface. A pattern of structured light incident on the animal surface is used to extract the surface geometry. A finite element based algorithm is applied to model photon propagation in the turbid media and a preconditioned conjugate gradient (PCG) method is used to solve the large linear system matrix. The reconstruction algorithm and the system performance are evaluated by phantom experiments.

We present a robust technique for diffuse optical fluorescence imaging of tumors in mice and tissue simulating fluorescence phantoms. The detection optics, which is a crucial part of a frequency domain fluorescence imaging system, with appropriate optical filters for efficient rejection of the excitation light, is demonstrated. The image reconstruction is divided into two parts; i.e. reconstructing the target locations such as size and position, and reconstructing the functional information such as fluorophore concentration and image reconstruction. The structural parameters i.e. tumor size and locations of the targets are recovered by a chi-square fitting technique by fitting the experimental data into analytically generated data. Having the structural information beforehand, the images are reconstructed by using our dual-mesh technique. The fluorescence images of targets of few tens of nanomolar fluorophore concentrations in both homogeneous and heterogeneous media are reconstructed in this study.

With the increasing importance of molecular imaging fluorescence based methods are continuously gaining impact. In fluorescence optical tomography excitation light is injected into the tissue where the fluorophore converts it to radiation of another wavelength. From the emitted light reaching the boundary the 3-D distribution of the fluorophore is reconstructed. This paper aims at finding the optimal spatial distribution of optodes in order to keep their number (hardware costs) low while gaining maximum information from the target object. The implemented algorithm starts with an arbitrary pool of feasible optodes. The optimal subset is searched by minimizing the mutual information between the different measurements. This goal is reached by subsequently removing those sources and detectors which add the least independent information until a stopping criterion is reached. Mutual information is estimated by calculating the inner products between the rows of the sensitivity matrix i.e. the first derivative of the forward mapping with respect to the optical parameters to be reconstructed. We assembled this matrix with a finite element implementation of the diffusion approximation of light propagation in scattering tissues. When starting with an initial pool of 96 optodes regularly spaced on a cylindrical surface and focusing on different target regions within the cylinder, the algorithm always converged towards physically reasonable optimal sets. Optimal source/detector patterns are be presented graphically and numerically.

We introduce a novel approach for localizing a plurality of discrete fluorescent inclusions embedded in a thick scattering medium using time-domain (TD) experimental data. It relies on numerical constant fraction discrimination (NCFD), a signal processing technique for extracting in a stable manner the arrival time of early photons emitted by one or many fluorescent inclusions from measured photons time of flight (TOF) distributions. Our experimental set-up allows multi-view TD data acquisition from multiple tomographic projections over 360 degrees without contact with the medium. Fluorescence time point-spread functions (FTPSFs) are acquired all around the medium with ultra-fast time-correlated single photon counting (TCSPC) after short pulse laser excitation. From these FTPSFs, the early photons arrival time (EPAT) of a fluorescent wavefront at a detector position is extracted with our NCFD technique. The key to our localization algorithm is to combine EPATs from several detection positions and projections to form 3D surfaces. The digital analysis of the concavities of the surfaces allows to find the 3D positions of an a priori unknown number of fluorescent inclusions located in the medium. Indocyanine green (ICG; absorption peak = 780nm, emission peak = 830nm) is used for the inclusions. Various experiments were conducted, and we show localization results on experimental data for up to 5 discrete inclusions distributed at arbitrary positions in the medium. We expect to extend our method to continuous distributions of fluorescence (rather than discrete inclusions) in a near future.

Hand-held based optical imagers have become a new research interest for its maximum patient comfort, less bulky instrument and potential for clinical translation towards breast cancer diagnostics. However, its ability for optical tomography is either limited by depth recovery since only reflectance measurements were obtained using a hand-held design for imaging. In this study, we introduced a self-guided multi-projection technique, which can take advantage of potential portability of hand-held probe based system, towards improvement of target depth recovery during fluorescence optical tomography studies.

We report our studies on the optical signals measured non-invasively on electrically stimulated peripheral nerves. The stimulation consists of the delivery of 0.1 ms current pulses, below the threshold for triggering any visible motion, to a peripheral nerve in human subjects (we have studied the sural nerve and the median nerve). In response to electrical stimulation, we observe an optical signal that peaks at about 100 ms post-stimulus, on a much longer time scale than the few milliseconds duration of the electrical response, or sensory nerve action potential (SNAP). While the 100 ms optical signal we measured is not a direct optical signature of neural activation, it is nevertheless indicative of a mediated response to neural activation. We argue that this may provide information useful for understanding the origin of the fast optical signal (also on a 100 ms time scale) that has been measured non-invasively in the brain in response to cerebral activation. Furthermore, the optical response to peripheral nerve activation may be developed into a diagnostic tool for peripheral neuropathies, as suggested by the delayed optical signals (average peak time: 230 ms) measured in patients with diabetic neuropathy with respect to normal subjects (average peak time: 160 ms).

We show the potential of functional near-infrared spectroscopy for the discrimination of mental workloads during a cognitive task with two different levels of difficulty. Standard data analysis based on filtering and folding average procedures were carried out to locate those source-detector pairs sensitive to the activated cortical regions. On these channels we applied two classification algorithms for the discrimination of mental workloads. Both algorithms showed a high percentage of successful classifications (>80%) on three over a total of four subjects where brain activation was detected. These results are comparable to standard scores found in the field of electroencephalography.

The development of diffuse optical tomography (DOT) methods for neuroimaging of humans is challenging due to the geometry and light level constraints. A high density imaging array system has been developed and used to demonstrate the possibility of true tomographic reconstruction of cortical activity within the adult subjects which are consistent with studies using functional MRI and positron-emission tomography. This work demonstrates the benefits of using high density imaging array by investigating depth related information available from the increased number of tomographic measurements. Through the use of depth related sensitivity analysis, it is shown that the use of 4th and 5th nearest neighbor (NN) measurements, the sensitivity of the data to absorption related changes within the brain are improved dramatically, as compared to 1st, 2nd or 3rd NN measurements. Additionally, it is shown that by the use of 5th NN measurements, it is possible to recover changes at depths of up to 20 mm within the brain, which is an improvement over the use of 4th NN.

The mean penetration depth of diffusely reflected photons is dependent on the arrival time t of photons, but not on the source-detector distance. Thus, all photons collected at the same t have the same depth sensitivity, and can be used for the reconstruction. Following this concept, we have implemented a system for 3D tomography using a single injection fiber and a time-gated ICCD camera. The feasibility of the novel approach to reconstruct a local perturbation was demonstrated both with simulations and phantom measurements. Finally, preliminary measurements were performed in vivo following a standard protocol of motor cortex activation.

This study examines differences in concentration changes of hemoglobin in the brain while finding algebraic solutions versus geometrical solutions. We use Near Infrared Spectroscopy imaging system to measure the hemoglobin changes while subjects are solving algebraic task and geometrical task. NIRS imaging system can measure changes in the concentration of hemoglobin. This brain activity data shows a difference between the two different experimental tasks which helps us to identify the characteristics of thinking processes.

Diffuse Optical Tomography (DOT) is based on acquiring information from multiply scattered light which penetrates into the tissue up to depths of several centimeters. This technique allows for imaging of absorbing and scattering inclusions inside tissue and distinguishing between them after computer processing of an image. An experimental setup for multicolor frequency-domain diffuse optical tomography (FD DOT) to visualize neoplasia of breast tissue and to estimate its size has been created. A breast is scanned in the transilluminative configuration by a single source and detector pair. Illumination at three wavelengths (684 nm, 794 nm, and 850 nm) which correspond to different parts of the absorption spectrum provides information about concentration of the main absorbers (oxygenated hemoglobin, deoxygenated hemoglobin, and fat/water). Source amplitude modulation at 140 MHz increases spatial resolution and provides separate reconstruction of scattering and absorption coefficients. In vivo study of breast carcinoma has been performed. Maps of 2D distributions of reconstructed absorption and scattering coefficients and concentration of hemoglobin have been obtained. An increase of absorption and scattering coefficient, total hemoglobin concentration and decrease of blood oxygen saturation is observed in the tumor area in comparison with the surrounding tissue. We can conclude that FD DOT technique confirms a possibility of detecting neoplastic changes.

We propose a 3D scheme for time-domain fluorescence molecular tomography within the normalized Born-ratio formulation. A finite element method solution to the Laplace transformed time-domain coupled diffusion equations is employed as the forward model, and the resultant linear inversions at two distinct transform-factors are solved with an algebraic reconstruction technique to separate fluorescent yield and lifetime images. By use of a multichannel time-correlation single photon counting system, we experimentally validate that the proposed scheme can achieve simultaneous reconstruction of the fluorescent yield and lifetime distributions with a reasonable accuracy.

This article aims at the development of the fast inverse Monte Carlo (MC) simulation for the reconstruction of optical properties (absorption coefficient and scattering coefficient ) of cylindrical tissue [1], such as a cervix, from the measurement of near infrared diffuse light on frequency domain. Frequency domain information (amplitude and phase) is extracted from the time domain MC with a modified method. To shorten the computation time in reconstruction of optical properties, efficient and fast forward MC has to be achieved. To do this, firstly, databases of the frequency-domain information under a range of μ_a and μ_s were pre-built by combining MC simulation with Lambert-Beer's law. Then, a double polynomial model was adopted to quickly obtain the frequency-domain information in any optical properties. Based on the fast forward MC, the optical properties can be quickly obtained in a nonlinear optimization scheme. Reconstruction resulting from simulated data showed that the developed inverse MC method has the advantages in both the reconstruction accuracy and computation time. The relative errors in reconstruction of the μ_a and μ_s are less than ±6% and ±12% respectively, while another coefficient (μ_a or μ_s) is in a fixed value. When both μ_a and μ_s are unknown, the relative errors in reconstruction of the reduced scattering coefficient and absorption coefficient are mainly less than ±10% in range of 45 < μ_s <80 cm^-1 and 0.25< μ_a <0.55 cm^-1. With the rapid reconstruction strategy developed in this article the computation time for reconstructing one set of the optical properties is less than 0.5 second.

Breast diffuse optical tomography is now highly expected as a potential routine inspection means for the high specificity and safety. Many efforts have been put to overcome its intrinsic adversities, such as low spatial resolution and quantitativeness. In this study, we propose a technique for enhancing image reconstruction of time-domain breast diffuse optical tomography (DOT). The technique uses finite-element-method (FEM) solution to the Laplace-transformed diffusion equation as the forward model, and an inverse model of Newton-Raphson iterative scheme. Through a target location that is provided by the preliminary image that is reconstructed using global optode arrangement or other techniques, we can obtain a more accurate image reconstruction by relocating all the optodes within the targeted region. The simulative experiments show that the performance of reconstructed image is evidently improved by the aid of the optodes relocation strategy.

Acquisition of the optical structures within a biological body is critical to all the diffuse light imaging modalities, such as diffuse optical tomography (DOT) and fluorescence molecular tomography (FMT). On an assumption of the optical homogeneity within the organs, it can be cast as a shape-based DOT issue, which aims at simultaneously reconstructing the boundary-describing parameters and optical properties of the disjoint domains of distinct tissue types. As the first step to the solution of this issue, we propose here a continuous-wave mode, elliptic-region-based DOT scheme. The methodology employs the boundary-element-method (BEM) solution to the diffusion equation as the forward model, and solves a nonlinear inverse issue that seeks an optimal boundary configuration as well as the optical properties to minimize the residual norm between measured and predicted data. The proposed scheme is validated using simulated data for a cylindrical geometry embedding two absorption- and scattering-contrasting ellipses at different noise levels.

Near infrared diffusion optical tomography (DOT) is one of promising tools for breast tumor detection because of its noninvasiveness and potential portability. In this paper, we propose an image reconstruction method for time-domain breast diffuse optical tomography, which offers simultaneous recovery of the absorption and scattering coefficients. Under the panel-compression detection mode, we propose a forward model based on the finite-difference solution to the diffusion equation, and furthermore develop an inverse model within a framework of the Newton-Raphson linearization and the generalized pulse spectrum technique. The proposed methodology is validated by experiments on a specifically-designed solid slab phantom containing two deeply-located absorption and scattering contrasting cylinders, using our multi-channel time-correlated single-photon-counting system.

In vivo trans-rectal near-infrared (NIR) optical tomography is conducted on a tumor-bearing canine prostate with the assistance of trans-rectal ultrasound (TRUS). The canine prostate tumor model is made possible by a unique round cell neoplasm of dogs, transmissible venereal tumor (TVT) that can be transferred from dog to dog regardless of histocompatibility. A characterized TVT cell line was homogenized and passed twice in subcutaneous tissue of NOD/SCID mice. Following the second passage, the tumor was recovered, homogenized and then inoculated by ultrasound guidance into the prostate gland of a healthy dog. The dog was then imaged with a combined trans-rectal NIR and TRUS imager using an integrated trans-rectal NIR/US applicator. The image was taken by NIR and US modalities concurrently, both in sagittal view. The trans-rectal NIR imager is a continuous-wave system that illuminates 7 source channels sequentially by a fiber switch to deliver sufficient light power to the relatively more absorbing prostate tissue and samples 7 detection channels simultaneously by a gated intensified high-resolution CCD camera. This work tests the feasibility of detecting prostate tumor by trans-rectal NIR optical tomography and the benefit of augmenting TRUS with trans-rectal NIR imaging.

Time-resolved polarization kinetics of near infrared spectral wing (SW) emission from human cancerous and normal prostate tissues was investigated. The intensity of spectral wing emission from the cancerous tissue was found to be stronger than that from the normal tissue, and the decay time of SW emission from the cancerous prostate tissue was found shorter than the normal prostate tissue. These differences can be attributed to increase of cell density during the evolution of tumor. The difference in intensities of the SW emission from cancerous and normal prostate tissues was used to image and identify cancerous areas from the surrounding normal prostate tissue.

Fluorescence Diffuse Optical Tomography is an optical non-invasive molecular technique for cancer imaging. Depending on the accessibility of the organ two main geometries might be considered, reflection or transmission. We will present first experimental and reconstruction comparison between these two geometries, on a laboratory time resolved bench. Both acquisitions were made using a fluorophore inclusion positioned in a liquid phantom, with breast comparable optical properties. We successfully reconstructed all fluorophore positions examined in both geometries. Reflection geometry suffers of many drawbacks that we have to deal with. We will present all challenges it implies, and also what are the advantages to use time resolved techniques in both geometries.

The purpose of this study was to extract scatter parameters related to tissue ultra-structures from freshly excised breast tissue and to assess whether evident changes in scatter across diagnostic categories is primarily influenced by variation in the composition of each tissues subtypes or by physical remodeling of the extra-cellular environment. Pathologists easily distinguish between epithelium, stroma and adipose tissues, so this classification was adopted for macroscopic subtype classification. Micro-sampling reflectance spectroscopy was used to characterize single-backscattered photons from fresh, excised tumors and normal reduction specimens with sub-millimeter resolution. Phase contrast microscopy (sub-micron resolution) was used to characterize forward-scattered light through frozen tissue from the DHMC Tissue Bank, representing normal, benign and malignant breast tissue, sectioned at 10 microns. The packing density and orientation of collagen fibers in the extracellular matrix (ECM) associated with invasive, normal and benign epithelium was evaluated using transmission electron microscopy (TEM). Regions of interest (ROIs) in the H&E stained tissues were identified for analysis, as outlined by a pathologist as the gold standard. We conclude that the scatter parameters associated with tumor specimens (N_patients=6, N_specimens=13) significantly differs from that of normal reductions (N_patients=6, N_specimens=10). Further, tissue subtypes may be identified by their scatter spectra at sub-micron resolution. Stromal tissue scatters significantly more than the epithelial cells embedded in its ECM and adipose tissue scatters much less. However, the scatter signature of the stroma at the sub-micron level is not particularly differentiating in terms of a diagnosis.

Over the past decade, the application of results from brain science research to education research has been a controversial topic. A NIRS imaging system shows images of Hb parameters in the brain. Measurements using NIRS are safe, easy and the equipment is portable, allowing subjects to tolerate longer research periods. The purpose of this research is to examine the characteristics of Hb using NIRS at the moment of understanding. We measured Hb in the prefrontal cortex of children while they were solving mathematical problems (tangram puzzles). As a result of the experiment, we were able to classify the children into three groups based on their solution methods. Hb continually increased in a group which could not develop a problem solving strategy for the tangram puzzles. Hb declined steadily for a group which was able to develop a strategy for the tangram puzzles. Hb was steady for a certain group that had already developed a strategy before solving the problems. Our experiments showed that the brain data from NIRS enables the visualization of children's mathematical solution processes.

Functional near-infrared spectroscopy (fNIRS) has the potential of easily detecting cerebral activity. However, in practical fNIRS measurements, a subject's physical or physiological changes such as body movements have often caused serious problems. If such a change is evoked by the tasks being monitored, it strongly correlates with the task sequence, and its interference in fNIRS cannot be eliminated using conventional signal filtering techniques. Hence, further improvement is necessary to eliminate such interference if we intend to use fNIRS on subjects with little or no physical restraint such as infants. We introduced an additional detector (d₂) between the source and detector (d₁) positioned in a conventional arrangement. The distances from the source to the detector d₁ and d₂ were set at 30 mm and 20 mm, respectively. Concentration changes of oxygenated and deoxygenated hemoglobin (ΔHbO and ΔHbR) were calculated using the linear combination of absorbance changes at d1 and d₂. Then tasks such as the upper-body tilting, the head nodding, the breath holding, and the finger opposition were performed by the participant. The statistical significance of the difference in concentration changes of ΔHbO and ΔHbR between task and rest periods was examined using the paired t-test. The results showed that interference due to upper-body tilting, head nodding, and breath holding was reduced by this method. Moreover, in the finger opposition task, a simultaneous increase of ΔHbO and decrease of ΔHbR was observed and these were significantly localized in the activation area by this method.

A three dimensional (3D) photon transport model has been developed based on the frequency domain simplified spherical harmonics approximation (SP_N) to the Radiative Transport Equation. Based on preliminary Monte Carlo studies, it is shown that for problems exhibiting strong absorption, the solutions using the 7th order SP_N model (N = 7) are significantly more accurate than those from a standard Diffusion (SP₁) based solver. This advance is of particular interest in the field of bioluminescent imaging where the peak emission of light emitting molecular markers are closer to the visible range (500 - 650 nm) corresponding to strong absorption due to hemoglobin.

In this work we studied the accuracy of a non-linear fitting procedure, based on the Levenberg-Marquardt algorithm, for time-resolved measurements to retrieve the absorption and the reduced scattering coefficients of an absorbing diffusive medium. This procedure is suitable for retrieving optical properties in a wider range of situations (e.g. solid samples, reflectance geometry), with respect to the linear inversion procedures recently presented for both CW and time domain measurements. By means of both analytical and numerical (Monte Carlo) simulations, we quantified the influence of photon counts, temporal sampling, analytical model, background and instrument response function on the accuracy in the estimation of the optical properties. Also a new analytical model to describe light propagation in diffusive media based on the Radiative Transport Equation has been considered. The main source of error that affects the accuracy of the absorption and reduced scattering coefficients retrieved by the non-linear procedure appears to be the analytical model adopted in the inversion procedure.

We introduce in this work a PDE-constrained approach to optical tomography that makes use of an all-atonce reduced Hessian Sequential Quadratic Programming (rSQP) scheme. The proposed scheme treats the forward and inverse variables independently, which makes it possible to update the radiation intensities and the optical coefficients simultaneously by solving the forward and inverse problems, all at once. We evaluate the performance of the proposed scheme with numerical and experimental data, and find that the rSQP scheme can reduce the computation time by a factor of 10 to 25, as compared to the commonly employed limited memory BFGS method.

The frequency-domain experimental data is typically corrupted by noise and the measurement accuracy is compromised. Assuming the widely used shot-noise model, it is well-known that the signal-to-noise ratio (SNR) of the amplitude signal decreases with increasing frequency, whereas the SNR of phase measurement reaches a peak value in the range between 400 MHz and 800 MHz in tissue volumes typical for small animal imaging studies. As a consequence, it can be assumed that there exists an optimal frequency for which the reconstruction accuracy would be best. To determine optimal frequencies for FDOT, we investigate here the frequency dependence of optical tomographic reconstruction results using the frequency-domain equation of radiative transfer. We present numerical and experimental studies with a focus on small tissue geometries as encountered in small animal imaging and imaging of human finger joints affected by arthritis. Best results were achieved in the 400-800 MHz frequency range, depending on the particular optical properties.

Multispectral and phase-contrast diffuse optical tomography are used to track treatment progress in a patient with locally advanced invasive carcinoma of the breast cancer during neoadjuvant chemotherapy. Two types of chemotherapy treatment including four cycles of Adriamycin/Cytoxin (AC cycles) and twelve cycles of Taxol/Herceptin (TH cycles) were applied to patient. A total of eight optical exams were performed before and within the chemotherapy. Images of tissue refractive index, and absorption and scattering coefficients, as well as oxy-hemoglobin and deoxy-hemoglobin concentrations along with scattering particle volume fraction and mean diameter of cellular components were all obtained. The tumor was identified through absorption and scattering images. Tumor shrinkage was observed during the course of chemotherapy from all the optical images. Our results show that oxy-hemoglobin, deoxy-hemoglobin and total hemoglobin in tumor decreased after chemotherapy compared to that of before chemotherapy. Significant changes in tumor refractive index along with tumor cellular morphology during the entire chemotherapy are also observed.

In this study, we present a simplified spherical approximated-radiative transport model-based reconstruction algorithm for three-dimensional diffuse optical tomography of finger joints. The reconstruction algorithm is tested and validated using a series of phantom experiments and in vivo data from healthy and osteoarthritic joints. In all the phantom and in vivo experiments, the absorption and scattering images are recovered using both the transport model and diffusion approximation based reconstruction algorithms. We observe that while both the diffusion and transport based algorithms can effectively image the finger joints, the transport model based can provide quantitatively improved reconstruction quality over the diffusion approximated based.

The fluorescence properties of reduced nicotinamide adenine dinucleotide (NADH) and oxidized flavoproteins (Fp) such as flavin adenine dinucleotide (FAD) in the respiratory chain are sensitive indicators of intracellular redox states and have been applied to the studies of mitochondrial function with energy-linked processes. The redox scanner, a threedimensional (3D) redox cryo-imager previously developed by Chance et al., can quantitatively determine the metabolic properties of tissue samples by acquiring the fluorescence images of NADH and Fp. The redox ratios, i.e., Fp/(Fp+NADH) and NADH/(Fp+NADH), obtained on the basis of relative signal intensity ratios, provide a sensitive index of steady-state of the mitochondrial metabolism that has been determined for a variety of biological tissues. This paper presents the further development of the instrument by establishing a calibration method to quantify the concentration of the fluorophores and facilitate the comparison of redox images obtained at different time or with different instrument functions. Calibration curves of both NADH and Fp have been obtained using snap-frozen standard references with NADH concentration ranging from 150-1400 μM and Fp from 80-720 μM. Snap-freeze tissue samples such as human breast tumors xenografted in mice, normal mouse pancreases and spleens were imaged. The NADH and Fp concentrations as well as the redox ratios in the tissue samples were quantified based on the adjacent solution standards of NADH and Fp. The obtained multi-slice redox images revealed high heterogeneity of the tissue samples which can be quantitatively interpreted.

A near-infrared (NIR) tomography system has been built to allow for imaging thick tissue at high frame rate. This tomography system uses a spectrally encoded source arrangement consisting of eight fibers coupled from temperature controlled single mode laser diode sources with about 1 nm spacing in their lasing wavelengths, having an overall spectrum confined to within 10 nm in the NIR region. Eight fiber-coupled, high-resolution, CCD based spectrometers were used to detect the intensities and decode their source origin locations. All detection CCDs were frame-synchronized using a computer controlled external TTL trigger circuit in order to preserve the temporal kinetics of the detected signals. A set of static heterogeneous phantom imaging was performed on a 64 mm thick resin phantom to verify the linearity and accuracy of the system and algorithm. Furthermore, to test the performance of this system at high frame rate, a dynamically varying absorption contrast study was realized by letting India ink diffuse into the phantom inclusion while continuously imaging it at 20 frames per second. The algorithm and the results from these phantom studies are presented. The 20 frames/second exposure rate and ability to image tissue beyond 60 mm thick makes this system perfect for potential clinical imaging of pulsatile hemodynamics in breast tumors.

Histologic information is often the ground truth against which imaging technology performance is measured. Typically, this information is limited, however, due to the need to excise tissue, stain it and have the tissue section manually reviewed. As a consequence, histologic models of actual tissues are difficult to acquire and are generally prohibitively expensive. Models and phantoms for imaging development, hence, have to be simple and reproducible for concordance between different groups developing the same imaging methods but may not reflect tissue structure. Here, we propose a route to histologic information that does not involve the use of human review nor does it require specialized dyes or stains. We combine mid-infrared Fourier transform infrared (FT-IR) spectroscopy with imaging to record data from tissue sections. Attendant numerical algorithms are used to convert the data to histologic information. Additionally, the biochemical nature of the recorded information can be used to generate contrast for other modalities. We propose that this histologic model and spectroscopic generation of contrast can serve as standard for testing and design aid for tomography and spectroscopy of tissues. We discuss here the biochemical and statistical issues involved in creating histologic models and demonstrate the use of the approach in generating optical coherence tomography (OCT) images of prostate tissue samples.

Current imaging modalities allow precise visualization of tumors but do not enable quantitative characterization of the tumor metabolic state. Such quantitative information would enhance our understanding of tumor progression and response to treatment, and to our overall understanding of tumor biology. To address this problem, we have developed a wide-field functional imaging (WiFI) instrument which combines two optical imaging modalities, spatially modulated imaging (MI) and laser speckle imaging (LSI). Our current WiFI imaging protocol consists of multispectral imaging in the near infrared (650-980 nm) spectrum, over a wide (7 cm × 5 cm) field of view. Using MI, the spatially-resolved reflectance of sinusoidal patterns projected onto the tissue is assessed, and optical properties of the tissue are estimated using a Monte Carlo model. From the spatial maps of local absorption and reduced scattering coefficients, tissue composition information is extracted in the form of oxy-, deoxy-, and total hemoglobin concentrations, and percentage of lipid and water. Using LSI, the reflectance of a 785 nm laser speckle pattern on the tissue is acquired and analyzed to compute maps of blood perfusion in the tissue. Tissue metabolism state is estimated from the values of blood perfusion, volume and oxygenation state. We currently are employing the WiFI instrument to study tumor development in a BRCA1/p53 deficient mice breast tumor model. The animals are monitored with WiFI during hyperoxic respiratory challenge. At present, four tumors have been measured with WiFI, and preliminary data suggest that tumor metabolic changes during hyperoxic respiratory challenge can be determined.