The system theory was developed from the combination of the operational calculus of N. Wiener and the transfer theory of K. Kuepfmueller. The system theory is the basic for the optical transfer function. With the introduction of a so- called transfer (or system) function, the question arose how to apply this theory to problems in optical tissue diagnostics.

In this report, we present a method to reduce the cross-talk problem in optical tomography. The method described is an extension of a previously reported perturbation formulation related to relative detector values, and employs a weight matrix scaling technique together with a constrained CGD method for imaging reconstruction. Results from numerical and experimental studies using DC measurement data demonstrate that the approach can effectively isolate absorption and scattering heterogeneities, even for complex combinations of perturbations in optical properties. The derive method is remarkably stable to errors originating from an insufficiently accurate estimate of properties of the reference medium.

Near infrared (NIR) light has become a powerful tool for non-invasive imaging of human brain activity. Many systems have been developed to capture the changes in regional brain blood flow and hemoglobin oxygenation, which occur in the human cortex in response to neural activity. We have developed a multi-channel reflectance imaging system, which can be used as a `mapping device' and also as a `multi-channel spectrophotometer'. In the present study, we visualized changes in the hemodynamics of the human occipital region in multiple ways. (1) Stimulating left and right primary visual cortex independently by showing sector shaped checkerboards sequentially over the contralateral visual field, resulted in corresponding changes in the hemodynamics observed by `mapping' measurement. (2) Simultaneous measurement of functional-MRI and NIR (changes in total hemoglobin) during visual stimulation showed good spatial and temporal correlation with each other. (3) Placing multiple channels densely over the occipital region demonstrated spatial patterns more precisely, and depth information was also acquired by placing each pair of illumination and detection fibers at various distances. These results indicate that optical method can provide data for 3D analysis of human brain functions.

Biological tissues are classified as turbid media to visible and near infrared lights. Photons propagating in tissue generally experience hundreds of scattering events before eventually being absorbed or escaping through boundaries. A lot of attention has been paid to exploiting diffusive photon density waves, which is, however, associated with difficulties in image reconstruction. Early arriving photons consist of a very small portion of photons reaching a detector. As they undergo forward scattering, their trajectories aggregate together to form a thin beam which probes limited volume. This makes it possible to implement single scanning scheme and to acquire more independent information for high quality images. This work is a preliminary theoretical study concerning the sensitivity and image resolution of a time-gating system. The Controlled Monte Carlo method was used for simulating migration of early arriving photons. Optical parameters and phantom dimensions were so chosen that they were applicable to breast cancer detection. Acquired edge-spread functions manifest the possibility of millimeter spatial resolution. Sensitivity analysis was derived from perturbations caused by targets of various contrasts and sizes, with respect to different temporal windows.

The head is inherently a heterogenous 3 dimensional (3D) domain and any useful image reconstruction algorithm in optical tomography should be based on the true shape and dimensions of the domain to be imaged. However due to the computational complexity of the problem, most reconstruction algorithms have been based on either 2 dimensional (2D) domains or simple 3D homogenous models. In this work we use a complex multi-layered model of the head to generate data in the presence of known changes of absorption within the brain. In order to evaluate the reconstruction of images from 3D data, we use either a 2D mesh based on the outline of a single measurement plane, or a 3D mesh based on the forward model itself. We show that the use of a 2D model for the reconstruction of images from a complex heterogenous 3D model is not successful for absolute imaging of internal optical property. In 3D, image reconstruction using meantime data only from a 3D domain is not successful when reconstructing for absorption only while keeping background scatter constant and homogenous. Image reconstruction in 3D is greatly improved when a-priori structural knowledge is used. Reconstruction of internal absorption only, from difference data (data after a change in absorption minus data before a change) works both in 2D and 3D, however in 2D, any depth information available from the data is lost.

In this work we present the first fully three-dimensional image reconstruction scheme for optical tomography that is based on the equation of radiative transfer. This scheme builds on the previously introduced concept of model-based iterative image reconstruction, in which a forward model provides prediction of detector readings, and a gradient-based updating scheme minimizes an objective function, which is defined as the difference between predicted and measured data. The forward model is solved by using an even-parity approach to reduce the time-independent radiative transfer equation to an elliptic self-adjoint equation of second order. This equation is discretized using a finite element method, in which we apply a preconditioned conjugate gradient method with a multigrid-based preconditioner to solve the arising linear algebraic system. The gradient of the objective function is found by employing an adjoint differentiation method to the forward solver. Initial tests on synthetic data have shown robustness and good convergence of the algorithm.

The photon diffusion approximation provides an accurate description for the propagation of light in a highly scattering medium such as most tissue. However, complications arise at the boundaries between such a highly scattering medium and a clear medium, as at the locations where the light enters and exits the medium. Several approximations have been developed to address the input of light from a laser through an optical fiber, and several others have been developed to address the collection of light by an optical fiber and detector. The principle of reciprocity suggests that source and detector fibers could be treated identically, resulting in a simplification of the diffusive optical model. Such a formulation agrees quite well with Monte-Carlo and other models. In the present work, we apply the technique to some different situations, including the important practical problem of source and detector fibers which are not aligned to be normal to the surface. Simple analytical results based on this model will be compared to Monte-Carlo results.

Dual interfering source systems have been shown to be more sensitive to the presence of absorbing inhomogeneities than single-source systems. They are used to localize accurately deep inhomogeneities with backprojection methods. However, there is not actually a method to perform indirect imaging with such techniques. In this work, a tomographic reconstruction method for an absorbing perturbation in highly scattering media and for phased array system is derived. The perturbation approach to solve the heterogeneous diffusion equation is used with the first order Rytov expansion. Analytical weight functions necessary to solve the inverse problem are obtained. Validity of the derived analytical expressions was checked with reconstructions from simulated data obtained with finite difference solutions of the diffusion equation and for a media mimicking breast tissue.

Theoretical models available for optical reflectance with small source-detector separations are limited. In this study we reported two modified diffusion models for steady-state, broadband (500 - 590 nm) reflectance with source-detector separations of 0.125 - 0.25 mm and 0.25 - 1.0 mm. For the former case, the major modifications are D equals 1/3(mu) _s, and z₀ equals 1/(mu) _s'; for the latter case, an additional modification is to multiply the separation by a factor of 2. Monte Carlo simulations were used to confirm the modified models at three different hemoglobin concentrations with relative errors of 12% or smaller. These models provide the possibility to extract local optical and physiological parameters of blood-perfused tissues from reflectance data measured with a needle-like probe.

Pulse oximetry (oxygen saturation monitoring) has markedly improved medical care in many fields, including anesthesiology, intensive care, and newborn intensive care. In obstetrics, fetal heart rate monitoring remains the standard for intrapartum assessment of fetal well being. Fetal oxygen saturation monitoring is a new technique currently under development. It is potentially superior to electronic fetal heart rate monitoring (cardiotocography) because it allows direct assessment of both fetal oxygen status and fetal tissue perfusion. Here we present the analysis for determining the most optimal wavelength selection for pulse oximetry. The wavelengths we chose as the most optimal are: the first in the range of 670-720nm and the second in the range of 825-925nm. Further we discuss the possible systematic errors during our measurements, and their contribution to the obtained saturation results.

Absolute dual-parameter image reconstruction in optical tomography (OT) is a nonlinear and ill-posed problem, requiring a model-based iterative approach and an accurate and fully three-dimensional light transport forward model. These factors make OT a computationally expensive problem, resulting in reconstruction times that range from several minutes to hours, which is not acceptable in most clinical applications. In order to reduce reconstruction times we propose a coarse-grain parallel implementation of the inverse algorithm using a shared-memory threaded approach. Unlike most conventional parallelization strategies which operate on the level of the linear algebra subsystem our approach uses a parallelization at the application level, thereby leaving the underlying linear matrix solution routines untouched. This allows to exploit the inherently parallel structure of the problem, while at the same time utilizing fast direct or efficiently preconditioned iterative serial linear solvers, which in most cases can not be efficiently parallelized. We compare the performance of the coarse-grain parallel approach with a low-level parallelization of the linear conjugate gradient solver and show that the proposed method achieves significant performance improvements, thereby bringing reconstruction times within a clinically acceptable range.

A method to measure the diffusion coefficient of homogeneous and layered media, based on multi-distance measurements of time-resolved reflectance is proposed. The diffusion coefficient is retrieved by fitting procedures on the logarithm of the ratio of two reflectances measured at two different source-receiver distances. The inversion procedure uses analytical models derived from the diffusion equation and it is not affected by the boundary conditions used. For layered media the inversion procedure retrieves the diffusion coefficient of the first layer. The proposed procedure is simpler than others usually employed and also provides a reliable criterion for retrieval of information on the layered structure of the diffusive medium.

A near-infrared optical tomograph has been developed to obtain 3-dimensional images of scattering phantoms with optical properties similar to those of biological tissues. This experimental setup uses a femtosecond laser and a synchroscan streak camera. Thanks to three stepping motors, the phantom to be imaged is scanned in a parallel-beam mode. The time resolved detection of scattered photons is performed in the transmission mode, for different phantom positions obtained by two translations and one rotation stages. Regarding the data processing, we have developed a nonlinear image reconstruction algorithm based on the Newton-Raphson iterative method. It executes absorption and scattering mapping on the basis of characteristic data extracted from the recorded temporal point spread function of light transmitted through the studied object, such as the mean flight time of photon, the related variance or the integral intensity. First, 2D images using absolute or differential imaging schemes have been obtained for different scattering cylindrical phantoms possessing one or two more absorbing and/or more scattering inclusions. Reconstructed images have been compared to simulated and expected values. The results demonstrate that this system is a reliable and valuable platform for research on time-resolved optical tomography.

We consider the inverse scattering problem for the diffusion equation with general boundary conditions. Computer simulations are used to illustrate our approach in model systems.

In this work we investigate the effect of a layer on Diffuse Optical Tomography of tissue. Such layers could be tissue structures (such as the skin or a fat-layer) or layers formed by compression plates. Our analysis uses an analytical forward model that is based on the angular spectrum representation of the propagating photon density wave in a diffuse medium. The inversion employs a standard perturbation expansion based on the Rytov approximation that is uses appropriate volume segmentation and solved using the algebraic reconstruction technique. The results demonstrate that the effect of biologically relevant multi-layer schemes can lead to significant reconstruction errors both in terms of quantification and positional certainty. The work is focussed on geometries and optical properties typical to the human breast, however the results are general and can apply to other tissue as well.

Simulation of tomography reconstruction of several absorbing macro-inhomogeneities in a strongly scattering cylindrical body was carried out not using any prior information on the structure of macro-inhomogeneities. The data of time-domain measurements were simulated using a diffusion equation solving by FEM. For the reconstruction procedure the method based on curvilinear photon average trajectories (PAT) was used. This method allows solving the inverse problem of optical tomography in terms of a set of linear equations with a sparse coefficient matrix. Special methods for sparse matrix processing permit to decrease sufficiently the computing time for image reconstruction. The comparison was carried out for the results of application of the several matrix solvers used in conventional CT (such as MART, NNLS, LSQR, and their combinations) to PAT method. In order to estimate the computing time and the quality of reconstruction, a comparison was carried out between the images restored using PAT method and using multi-iterative program MMTOAST10, which employs FEM solving of the diffusion equation. It was shown that the trajectory method, being much faster, gives the comparable quality of the restored images.

In this paper, we describe the use of the full sixteen element Mueller matrix to differentiate between surface lesions on the skin, as an eventual noninvasive diagnostic technique for detection of certain types of skin cancers. A semi-automated system using a polarized light beam and a computer controlled CCD camera was developed to collect 16 polarization images of a sample and to calculate the complete Mueller matrix in near real time. The system was initially calibrated and the accuracy and precision were evaluated with a mirror and polarizer. This study also includes in vivo results from normal skin, a benign nevus, and a known cancerous lesion acquired from a single Sinclair swine. Differences were observed between the Mueller matrices of all three in vivo cases. These preliminary results demonstrate the potential for using an automated polarization imaging apparatus for eventual clinical cancer diagnostics.

We examined the effect of temperature change on the diffuse reflectance of the skin. The optical probe consists of several optical fibers located at the center of a thermal electric device, which controls the temperature at the surface of the skin in contact. Measured light reflectance profile between 0.4-1.9 mm was fitted to a mathematical model obtained by Monte Carlo simulation, and absorption and scattering coefficients were estimated. The reduced scattering coefficient of the forearms consistently showed a positive relationship with temperature between 22 and 42 degree(s)C. This dependency was reversible without apparent delay. The same effect was observed on ex vivo pigskin. It is possible to explain the positive instantaneous dependency of scattering on temperature by the change of the refractive index of intercellular fluid. The scattering coefficient of the subcutaneous fat of pigskin showed a negative dependence on temperature. This negative dependency of scattering can be attributed to a phase change as a function of temperature. The absorption coefficient in vivo also increased with temperature from 22 to 42 degree(s)C. But the change was not immediately reversible after temperature reached 40 degree(s)C. This relationship was similar to the nonlinear increase in blood perfusion observed in laser Doppler measurements.

Near infrared (NIR) spectroscopy and imaging have been applied to measure oxygenation change in the brain. Multi- channel NIR system can obtain topographic image of brain function. However poor spatial resolution due to tissue scattering is serious problem of NIR imaging. In this study, we propose a simple method to estimate the center of the absorption change in the brain from the signal detected with multi-channel NIR system. The theoretical analysis of photon measurement density functions for the adjoining source- detector pairs in multi-channel NIR system suggests that the difference in intensity between the adjoining source- detector pairs almost linearly relates with the position of the absorption perturbation. The proposed method to estimate the center of the absorption change in the brain is evaluated by using three layered head model. The estimated position is almost consistent with the center of the absorption change. In case where the broad absorption change occurred close to the edge of the measuring region, the estimated position tends to be shifted inward of the measuring region.

Diffuse optical tomography (DOT) can be considered as an optimization problem, in which the minimum of an objective function is sought. The objective function is typically some measure of the difference between the predicted and experimentally obtained detector readings. Most of the optimization techniques that are currently applied in optical tomography employ so-called gradient methods. These methods start from an initial guess of the distribution of optical properties and iteratively update this initial guess along the gradient of the objective function. It is well known that the success of gradient techniques depends strongly on the initial guess. If the guess is not chosen appropriately, the algorithm may not converge or may converge to a local minimum. Evolution strategies are global optimization techniques that depend much less on initial guesses. The drawback of evolution-based codes is that they are computationally expensive. In this work we introduce a hybrid approach that combines the advantages of gradient techniques and evolution strategies. The hybrid algorithm is less dependent on an initial guess and overcomes the computational burden connected to evolution strategies.

Vascular disease is a significant source of mortality and morbidity for many patient populations. While substantial strides in surgical therapeutics have been made in the past decade, our limited understanding of the microvascular processes, which are invisible to conventional imaging modalities and beyond the scope of our current physiologic paradigms, has slowed the advancement of medical therapeutic interventions. In this report we present data in support of an emerging body of work demonstrating that the method of dynamic optical tomography can yield critical insights into the underpinnings of microvascular pathophysiology in large tissue structures.

Functional imaging of muscle oxygenation using NIRS is a promising technique for evaluation of the heterogeneity of muscle function and diagnosis of peripheral vascular disease or muscle injury. We have developed a 200-channel imaging system that can measure the changes in oxygenation and blood volume of muscles and that covers wider area than previously reported systems. Our system consisted of 40 probes, a multiplexer for switching signals to and from the probes, and a personal computer for obtaining images. In each probe, one two-wavelength LED (770 and 830 nm) and five photodiodes were mounted on a flexible substrate. In order to eliminate the influence of a subcutaneous fat layer, a correction method, which we previously developed, was also used in imaging. Thus, quantitative changes in concentrations of oxy- and deoxy-hemoglobin were obtained. Temporal resolution was 1.5 s and spatial resolution was about 20 mm, depending on probe separations. Exercise tests (isometric contraction of 50% MVC) on the thigh with and without arterial occlusion were conducted, and changes in muscle oxygenation were imaged using the developed system. Results showed that the heterogeneity of deoxygenation and reoxygenation during exercise and recovery periods, respectively, were clearly observed. These results suggest that optical imaging of dynamic change in muscle oxygenation using NIRS would be useful not only for basic physiological studies but also for clinical applications with respect to muscle functions.

The utility of optical tomography as a static imaging modality is limited by its intrinsically low spatial resolution and quantitative accuracy. When applied to dynamic measurements, however, optical imaging methods have the potential to assess tissue function as revealed by temporal variations in tissue optical properties. These variations are a consequence of vascular hemodynamic processes, which are known to exhibit considerably spatiotemporal heterogeneity. In this report we provide evidence, from simulation, that complex dynamic behavior in optical coefficients occurring in localized regions in highly scattering media can be accurately characterized by the method of dynamic optical tomography, even in the limiting case of spatiotemporally coincident behavior.

We describe the design rationale, performance features, and operating characteristics of a newly constructed CW-NIR tomographic imaging system that is capable of continuous, real-time imaging of large tissue structures. Results from phantom and clinical studies are presented and discussed.

A discussion of photon migration will introduce both some fundamental limitations and practical concerns which directly affect the design and construction of DOT instrumentation. The utility of four different multiplexing schemes will be discussed and then contrasted in light of these concerns. The relative merits of each scheme will be used to determine which measurements they are best suited for, given some of the realistic technical, legal, and practical issues common to the medical research community. Within this context, the unique advantages offered by a variant of time-division multiplexing; Pulse-TDM, will be introduced and explained.

In this paper we present a system for optical imaging in turbid media based on an intensified CCD camera. The time-gating capability of the light intensifier has been used to acquire time sliced images of the light transmitted through phantoms made of intralipid and ink solutions. Then, we analyzed the images by fitting the time series with a theoretical expression for photon migration. This allowed us to recover the maps of the absorption ((mu) _a) and scattering ((mu) _s') coefficients in an extended area of the samples. Using our system a possible heterogeneity in the sample can be easily detected. We characterized the performances of the system following two approaches. First, we evaluated the linearity of the relationship between the recovered absorption coefficient and the ink concentration and between the recovered scattering coefficient and the intralipid content (homogeneous samples). Then, we measured the apparent contrast in the (mu) _s' maps resulting from small inclusions with a scattering coefficient higher and lower with respect to a homogeneous background (heterogeneous samples). Results encourage further researches.

A preliminary study of the effect of a spatially heterogeneous background upon near-infrared tomographic imaging is presented using computer simulations and phantom experiments. In simulations, increasing heterogeneity is added to a sample region and the reconstructed images are analyzed. A finite-element reconstruction algorithm based on the frequency-domain diffusion equation is used. A phantom study for a heterogeneous object was performed by embedding an object in excised breast tissue. The ability to distinguish and quantify a heterogeneous inclusion in a heterogeneous background is discussed. In this case, a 20 mm diameter object in a 100 mm background is characterized with up to 75% added heterogeneity in simulation. A 30 mm cube of pork embedded in a 86 mm sample of excised breast tissue is localized and recovered to within 25% of its expected contrast.

Previous studies have suggested that the phased-array detection could achieve high sensitivity in detecting and localizing inhomogeneity embedded within turbid media by applying dual interfering sources. While the sensitivity of the phased array system will be influenced by various factors such as the source-detector configuration, the effect of noises (variation of source strength and phase, detector noise, imperfect homogeneous medium, etc.) and the modulation frequency. Analytical solutions as well as experimental data are presented to investigate the sensitivity of phased array detection using different degrees of absorption perturbation by varying the size and contrast of the absorber. Also we compare the result with the single source case under similar configuration. Different separations between two interfering sources have been compared to yield the maximum sensitivity for a typical perturbation. These results can be helpful for modifying the experimental apparatus to reach optimal detection sensitivity.

Continuous-Wave (CW) and Frequency-Domain (RF) methods of diffuse optical imaging are compared to identify the optimal approach for a clinical system. The CW Imager features fast imaging with 9 frequency-encoded sources and 16 detectors. The homodyne RF imager has 40 sources modulated at a frequency of 70 MHz and 9 detectors. Both systems incorporate two wavelengths. The features and characterization of both system are discussed and compared. A versatile phantom box design is presented that offered flexibility in experimental design and facilitated optimization of probe geometry. A 3D absorption image reconstruction phantom experiment is carried out to test the performance of the RF system. Noise models are set up and experimentally verified, and a signal-to-noise ratio analysis is presented. The heterogeneity detection sensitivity of the CW method and the RF method are compared, using both simulation and phantom experiments. The comparison results are discussed in the context of the most suitable approach to the design of a clinical breast cancer detection system.

A generic analysis for determining the accuracy of extracting the optical coefficients from a scattering medium is presented for measurement systems including CW, temporal and frequency domain configurations. The analysis is based on the relationship between the optical coefficients and a pair of measurables from a measurement system enabling the efficacy of a system to be gauged with the use of analytical models or Monte Carlo simulations. The methodology is sufficiently versatile to allow the performance of different types of measurement system to be compared. For each system considered, different measurables are evaluated to obtain the optimal accuracy. Examples of the many suitable measurables utilized include the moments of the detected intensity distribution in the CW and temporal domain configurations and the phase and modulation of a detected signal in the frequency domain. The error in reconstructing the absorption coefficient from simulated and analytical scattered data is calculated for the frequency and time domain cases and a CW case where the intensity is measured as a function of the distance across a slab of scattering material. The results indicate that cost effective techniques based on measuring the CW intensity have promise in situations where the thickness of the scattering medium may be selected.

This paper summarizes the design steps that are followed during the development of the portable optical imager for breast cancer screening. The design steps considered the parameters such as total power consumption versus battery weight and size, speed of data acquisition versus cost and complexity of the design (functionality), graphical display versus operating system choice. We have used a single board computer system that uses Windows CE as the real time operating system. This choice was preferred since our graphical display requirements can only be carried out with the CE environment's GUI kernels.

We report on the first three dimensional tomographic localization of vascular reactivity in the brain. Using a model-based iterative image reconstruction algorithm we show volumetric spatial changes in the absorption coefficient caused by changes in blood volume. Unlike currently available topographic reconstruction techniques, volumetric reconstruction schemes promise to be capable of spatially distinguishing between signals originating in the cerebral cortex from those originating in the overlying vascular tissues.

The quantification of tissue properties by optical measurements will facilitate the development of noninvasive methods of cancer diagnosis and detection. Optical measurements are sensitive to tissue structure which is known to change during tumorigenesis. The goals of the work presented in this paper were to verify that the primary scatterers of light in cells are structures much smaller than the nucleus and then to develop an optical technique that can quantify parameters of structures the same size as the scattering features in cells. Polarized, elastic back-scattering was found to be able to quantify changes in scattering properties for turbid media consisting of scatterers of the size found in tissue.

Measurements of optical properties of biological tissues are necessary to develop optical diagnostic instruments. Although various methods to measure the optical properties of biological tissue have been proposed, very few measurements have been carried out in vivo because of their difficulties. We are developing a new method to obtain the optical properties in vivo using time-resolved measurement system. This new method determines the optical properties by comparing the experimental results of time-resolved measurement with the results of a series of Monte Carlo simulation using known optical properties.

In this work time-resolved reflectance spectroscopy (TRS) measurements were performed in vivo on biological tissues (arm and forehead) over a broad spectral range (610-1000 nm) and at different interfiber distances (2-4 cm), by a fully automated system based on mode-locked, continuously tunable laser sources and on time-correlated single photon counting for detection. Preliminary estimates of average tissue optical properties are obtained by interpreting the TRS curves at each wavelength with a homogeneous model for photon migration. The scattering coefficient decreases progressively upon increasing wavelength, while the absorption line shapes show the typical spectral features of the principal tissue components (hemoglobin, water, and lipids), with different weight depending on the tissue type and structure. The best fit of the absorption spectra with the spectra of the pure constituents and of the reduced scattering spectra to Mie theory provided information on the average percentage composition of the different tissues and on the average equivalent tissue structure, respectively. Moreover, TRS measurements at different interfiber distances allowed us to probe different tissue depths, showing the presence of a inhomogeneous structure with different average optical properties.

Tracking a bolus of contrast agent traveling through the cerebral vasculature provides a measure of the blood supply and the blood flow velocity in the respective cerebral tissue. This principle has been the basis for the first approaches in functional MR imaging and is of great value when investigating stroke and other vascularly compromised patients. While the bolus measurement is a standard procedure in clinical MR imaging, optical bolus tracking has not yet become a reliable protocol. Here optical absorption changes induced by bolus signals of the dye indocyanine- green are studied by near infrared spectroscopy on volunteers. The aim is to assess the latency and shape of the absorption change.

This is the report on the use of multichannel near-infrared spectroscopy (MNIRS) with indocyanine green (ICG) to determine regional cerebral blood flow (rCBF) distribution disturbance in infants. We measured rCBF in an infant with subdural hemorrhage after surgical removal of a subdural hematoma. A probe consisting of 12 optical fibers, 6 for transmission and 6 for detection, was set on the right and then left temporal regions of the head of the infant, and 16 measuring points were determined. Changes in ICG concentration were recorded using MNIRS (near infrared optical imaging system, OMM-2000, Shimadzu Corp., Japan).

A clinical study of the evolution of the space resolved reflectance with the subcutaneous layers thickness, at rest, on the vastus lateralis, is used to determine convenient sets of muscle and fat optical coefficients at 850 nm. The results are compared to the data available in the literature. When the same study will be completed at 780 nm, these coefficients will be used in 3 layers MC simulations to establish pre-calculated tables of the reflectance as a function of the fat thickness and of the muscle absorption coefficient. The goal of this study is eventually to perform real time inversion during effort in order to determine the muscle oxygenation with this very simple technique (part 2).

We examine the ability to recover the optical properties of a two-layer turbid medium using multi-distance frequency domain reflectance measurements and a hybrid Monte Carlo-- diffusion model. Frequency domain measurements are performed on two-layer liquid tissue phantoms simulating skin on muscle and skin on fat. Particular care to systematic effects in the photomultiplier is required for the measurements at short source-detectors distances. The model converges when fitting five free parameters (the optical properties of the upper and lower layers and the upper layer thickness). However, discrepancies between experimental and model yield insufficient accuracy for the absorption coefficient of the upper layer.

We study the correlation between (mu) _s' and THC obtained in vitro, in a highly scattering medium containing human blood. We used a frequency domain near infrared spectrometer (modulation frequency: 110 MHz, wavelengths: 758 and 830 nm) to measure in real time (acquisition time: 0.64 s) (mu) _s' and THC. We used Liposyn suspension and red blood cells in saline buffer solution. After a couple of minutes of baseline acquisition, several consecutive increments of 3-5 ml blood were added to the solution yielding THC equals 15-100 (mu) M and (mu) _a equals 0.03-0.3 1/cm. At the last amount of blood added, increments of glucose in the range of 0.5-20 g/L were added. For each step of blood and glucose added, data were acquired for a couple of minutes. This was repeated 6 times. Average of data was calculated for both (mu) _s' and THC for each of the red blood cells and glucose increments added. We found a high correlation between (mu) _s' and THC (0.018 X THC + 4.51, R² equals 0.98 at 758 nm and 0.012 X THC + 4.86, R² equals 0.97 at 830 nm). We studied the effect of glucose on (mu) _s' and we found a high correlation between the glucose added to the suspension and the decrease in (mu) _s' for the case of high glucose concentrations. The slope of this correlation is -0.011 at both wavelengths and the correlation factors were R² X 0.96 at 830 nm and R² equals 0.91 at 758 nm (case shown). The effect of glucose was less significant at 830 nm than at 758 nm in general. This work is a proof of principle for detection of (mu) _s' changes with glucose. This approach also establishes limits for glucose detection in physiological conditions.

We propose the use of near-infrared spectroscopy (NIRS) to monitor neonatal brain temperature non-invasively and continuously during hypothermic treatment by measuring changes in the NIR water absorption spectrum. Accurate measurements of the water spectrum in the range 650-1050 nm between 42 and 32 degree(s)C have been made and calibrated using a principal component regression, giving a standard error of calibration of <0.1 degree(s)C for 3 or more principal components. In vivo attenuation spectra measured on the adult forearm as a function of temperature have been employed in a preliminary investigation. In order to predict tissue temperature from changes in the water spectrum, we must first account for the effects of light scattering and haemoglobin absorption. A non-linear difference technique based on diffusion theory has been used to reconstruct the total absorption for each tissue spectrum, with respect to an assumed initial state. Second derivative spectroscopy was used to estimate the initial tissue chromophore concentrations for the reference spectrum. A linear fit to the reconstructed absorption spectra was then performed, using the NIR chromophore spectra and the water loading vectors determined from the PCR calibration. Tissue temperature was predicted with a standard error of approximately 1 degree(s)C.

We have developed a phase and intensity-modulated spectroscopy system (PMS) using a newly developed algorithm based on the microscopic Beer-Lambert law. Experiments with phantoms and the human body demonstrate the feasibility and reliability of the system as well as the new algorithm. Our goal is to develop compact, cost-effective, highly reliable, and user-friendly medical equipment for the quantitative monitoring of oxygen metabolism, and so on. The PMS system consists of three time-shared wavelength laser diodes with a 70MHz modulation frequency as sources and a 3mm diameter silicon PIN photodiode as a detector with an in-phase quadrature demodulator (IQD) for AC amplitude and phase detection. The PIN photodiode is operated at a low voltage and is durable against strong extraneous light. In addition, a specially designed low-noise amplifier is achieve a high S/N and reliable measurement. Our algorithm is independent of boundary conditions, exterior shape, scattering properties of the medium, and optode separation for measurement. We can therefore quantify the absolute concentration for oxy- deoxy-hemoglobin and hemoglobin saturation in living tissue of various shapes precisely.

Diffuse optical tomographic imaging has broad application to the field of clinical diagnosis of diseased tissue states. In particular, there has been considerable interest in applying diffuse optical imaging to breast cancer detection. We explore the effects of such parameters as regularization, source-to-source and detector-to-detector spacing, voxel size and source modulation frequency on characteristics of the point spread function (PSF) of the reconstructed image. An understanding of these effects is critical to the optimization of diffuse optical tomographic imaging techniques. We also present a figure of merit that can be used to compare various reconstruction techniques and parameters when considering the effects of noise on quantitative accuracy and resolution. In all cases we look at heterogeneities with changes in absorption only and limit the analysis to the linear approximation of the diffusion equation. We show that the resolution and quantitative accuracy are fundamentally limited by the degree of regularization required to achieve images of acceptable contrast in the presence of representative noise levels. All simulations are based on Tikhonov regularization and truncated singular value decomposition (TSVD) reconstruction techniques for which we have generally achieved best results.

A technique for measuring broadband near-infrared absorption spectra of turbid media is presented using a combination of frequency-domain (FD) and steady-state (SS) reflectance methods. Most of the wavelength coverage is provided by a white-light SS measurement, while the FD data are acquired at a few selected wavelengths. Coefficients of absorption ((mu) _a) and reduced scattering ((mu) _s') derived from the FD data are used to intensity-calibrate the SS measurements and to estimate (mu) _s' at all wavelengths in the spectral window of interest. After these steps are performed, (mu) _a can be determined by comparing the SS reflectance values to the predictions of diffusion theory, wavelength by wavelength. We present an application of this method to breast tumor characterization. A case study of a fibroadenoma is shown, where different absorption spectra were found between the normal and the tumor sides.

Image reconstruction and data collection in optical tomography can be achieved in a number of different ways. This paper explores the limitations of using assumptions of linearity, particularly in the case where image data is acquired before and after a change in optical properties within an object with heterogeneous optical properties. The effects of using a 2 dimensional (2D) reconstruction scheme for changes in 3D measurements are also demonstrated. Problems are a direct result of the inherent non-linearity of optical tomographic image reconstruction. We show how these assumptions affect images of changes in absorption in the presence of a) heterogeneous background scatter, and b) heterogeneous background absorption using both simulations and time-resolved experimental data. Comparisons of results using non-linear and linear image reconstruction techniques are included throughout. The origin and dependence of the error are investigated. Methods to improve results by using estimates of background structure from baseline images are shown to improve quantitation and object localization in simple images. The potential significance of this error is discussed in relation to successful, reliable clinical imaging of the neonatal brain.

Authors proposed optical topography to visualize the blood- volume change in the cortex associated with brain activation. The diffusion equation was used to obtain the sensitivity distribution of the blood-volume change. We show that the sensitivity distribution agrees with that obtained by the Monte-Carlo simulation. We then developed a phantom that simulates the light scattering property and brain activation in the cortex. Topographic images of the absorber in the phantom are obtained, and high location accuracy but spatial resolution of the topographic image were found. In addition, we derived a methodology to arrange optodes at high density in order to improve the spatial resolution of the topographic image.

An optical tomographic imaging system that is composed of 32 near infrared time-resolved spectroscopic channels has been newly implemented for the experimental evaluation of diffuse optical imaging schemes. This article reports on some initial results reconstructed from the transmitted measurement of such a time-resolved system. Two cylindrical tissue-like phantoms having two- and three-dimensional interior structures respectively are employed for the validation of the reconstruction algorithm, which uses the diffusion-based photon migration model and an optimization-based reconstruction solver. We compare the performances of the proposed two- and semi three-dimensional algorithms in reconstructing two-dimensional phantom and demonstrate the ability of full three-dimensional algorithm to simultaneously image multiple slices of there-dimensional phantom, using absolute and/or differential imaging schemes. Also addressed are some technical issues on the calibrations of the time-resolved measurements for matching the photon migration model and alleviating the influence of the inconsistency of the detecting channels. Our reconstructions show that the image quality can be significantly improved by differential imaging scheme that takes radio or difference of the real and reference measurement as the data type. Moreover, the results demonstrate the feasibility of experimental reconstruction from absolute mean time of flight of the temporal measurement, and argue that finer adjustments for the consistency of the detecting channels are necessary to improve the image quality of the absolute reconstruction.

The monitoring of a single muscle location does not reflect the heterogeneity of the muscle groups activation during exercise. In the past, measurements of oxygen consumption (VO₂) at single muscle locations could be carried out non-invasively by near-infrared continuous wave spectroscopy (NIR_CWS) at rest or during isometric contractions. In the present study, human regional quadriceps (vastus lateralis and rectus femoris) VO₂ was investigated at rest and during maximal voluntary contractions using a 12- channel NIR_CWS system with an acquisition time of 0.1 s.

This paper describes the theory and the apparatus on our newly developed 64 channel optical imager that enables imaging of three components, indocyanine green (ICG), oxyHb and deoxyHb. Knowledge on blood flow is obtainable using ICG as a blood flow indicator that is to be introduced via injection. ICG has a strong absorption at 805nm, therefore the changes in ICG is detectable with NIR spectroscopy separating from oxyHb and deoxyHb using proper three wavelengths. In addition to the system description, it describes an animal experiment and a phantom test for the validation study. It discusses independence features of ICG on Hb values. Using a simple quasi Lambert Beer's law, it proved the excellent separation of the three components. It also discusses the factors that affect the performance of signal separation.

The aim of this study was to compare functional cerebral hemodynamic signals obtained simultaneously by near infrared spectroscopy (NIRS) and by functional magnetic resonance imaging (fMRI). The contribution of superficial layers (skin and skull) to the NIRS signal was also assessed.

Time-resolved measurement enables one to restrict the volume of tissue sampled by NIRS instrument. The light propagation in the real head is more complicated, especially early detected photons are affected by the presence of low scattering cerebrospinal fluid surrounding the brain. In this study, time-resolved spatial sensitivity profiles in the realistic adult and neonatal head models are predicted by Monte Carlo simulation. The geometry of the head models is based upon the MRI scan of the adult and neonatal heads. The photon propagation is predicted by the delta-scattering technique of Monte Carlo simulation. The time-resolved spatial sensitivity profiles in the adult head are strongly affected by the low scattering CSF layer while the effect of the CSF in the neonatal head model is less significant than that in the adult head model.

Optimization of the systems and reconstruction techniques is important for good optical breast imaging. Quantitative and objective assessments of the quality of the reconstructed image and the detection sensitivity are required to optimize single projection optical imaging. In this paper, the detection sensitivity was estimated by the contrast to noise ratio, and the quality of an image is evaluated using resolution, image noise, positional error, and quantitative accuracy. The expressions of these evaluating parameters were set up and analyzed by simulation. Using the quantitative assessment parameters we developed, we present a sensitivity and accuracy analysis of the Truncated Singular Value Decomposition (TSVD) method applied to optical breast imaging in transmission mode. The results indicate that the higher spatial resolution and improved quantitative accuracy obtained by decreasing the TSVD regularization parameter (i.e. the number of singular values used in the matrix inversion) must be balanced against the concomitant increase in image noise. Other parameters needs to be optimized including probe geometry, voxel size were also analyzed.

A pilot clinical study on ten female volunteers is reported using our multi-channel frequency-domain optical imager. Seven of these patients were previously identified with tumors or microcalcifications by x-ray and/or ultrasound mammography. These tumors had a size between 5 and 38 mm. Using our optical imager and reconstruction methods we were able to detect all tumors that had been identified by x-ray and/or ultrasound mammography. The quantification of our improved scattering images shows a clear differentiation between benign and malignant tumors. A comparison between our results and x-ray/ultrasound results is also given.

Near-infrared imaging is being examined clinically as a diagnostic tool to image hemoglobin concentration and oxygen saturation within tissue. Optical projection measurements can be detected from tissue surfaces and combined with an appropriately designed diffusion model-based image reconstruction algorithm to recover maps of the tissue interior. One important issue in the development of this modality is that the geometry of source and detector fibers has a significant impact of the resulting images. In this paper we examine the tradeoffs between using a full circular array of measurements versus a partial array of measurements, as often described in the fan beam imaging approach. Case examples are presented to qualitatively illustrate the possible improvement in going from a circular geometry to the fan-beam geometry of detectors.

Near infrared diffuse optical spectroscopy and imaging may enhance existing technologies for breast cancer screening, diagnosis, and treatment. NIR spectroscopy yields quantitative functional information that cannot be obtained with other non-invasive radiological techniques. In this study we focused upon the origins of this contrast in healthy breast, especially from water and lipids.

The effective measurement of dynamic changes of blood and tissue oxygenation of tumors could be valuable for optimizing tumor treatment plans. For this study, a near- infrared spectroscopy system and pO₂ needle electrode were used to measure simultaneously changes in total hemoglobin concentration ([Hb]_total), oxygenated hemoglobin concentration ([HbO₂[) and local oxygen tension (pO₂) in the vascular bed of prostate tumors implanted in rats in response to respiratory challenge. The inhaled gas was alternated between air and carbogen (95% oxygen, 5% CO₂). Significant changes in tumor vascular oxygenation were observed with an apparent threshold for variation in [HbO₂]/[HbO₂]_max. For comparison, a phantom study was undertaken with 1% intralipid solution and blood. The slope of [HbO₂]/[HbO₂[_max vs. pO₂ in the phantom was ten times larger than in the tumor indicating that tumor cells are relatively resistant to oxygenation. This study demonstrates that the NIR technology can provide an efficient, real-time, non-invasive approach to monitoring tumor physiology and is compatible with additional techniques.

In this paper, we demonstrated a novel method to utilize a scattering near infrared light localized by an amplitude-modulated ultrasonic wave to image a buried object in a dense scattering medium through the synchronous scanning of the focused-ultrasound and a NIR laser beam. The frequency of the ultrasound (carrier wave) for localization of scattering light is 1MHz, and its amplitude is modulated by a low frequency sine waveform (modulation wave, 10KHz). The low frequency light signal from the focal zone of the ultrasound demodulated by real-time FFT. This scheme can increase a signal-to-noise ratio (SNR) of the ultrasound-localized-optical-tomography of tissues structure with different optical parameters. A better-quality optical tomography image of buried objects within simulating tissue is reconstructed by the detection and demodulation of the scattering light under the action of amplitude-modulated ultrasonic wave.

A Newton-Raphson inversion algorithm is extended for simultaneous absorption and scattering reconstruction of full 3-dimensional (3D) time-resolved diffuse optical tomography, which is based on efficient computation of Jacobian matrix of the forward model, and uses Algebraic Reconstruction Technique as the linear inversion tool. We validate the algorithm using both numerically simulated data from 3D finite-element discretization model of a tissue-like cylindrical phantom, embedding three inhomogeneities with different geometric and optical properties, as well as experimental measurements for a 3-D structural cylindrical phantom containing three absorptive rods, by means of fan-beam scanning mode. Our results show that a full 3D simultaneous reconstruction of absorption and scattering in turbid medium can be achieved with highly reasonable quality provided that the diffusive nature of light propagation in tissue is appropriately modeled, and temporal information of the measurements effectively employed.

The accurate forward modeling of light propagation in tissue to estimate the partial optical path length in the brain is important for quantitative optical spectroscopy and optical imaging. In this study design of simplified adult head model is investigated to estimate the partial optical path length in the brain. We consider homogeneous, two and three layered models as the simplified models. The apparent optical properties of surface layer of the simplified model are determined by TPSFs at small source-detector separations. Since it is difficult to determine the optical properties of the brain from the in-vivo time-resolved measurement, the optical properties of the brain are assumed to be the reported data. The proposed modeling is evaluated by Monte Carlo simulation. The TPSFs of the five layered models that consist of the scalp, skull, CSF, grey matter and white matter are predicted by Monte Carlo simulation. The apparent optical properties of the surface layer are estimated from the TPSFs for five layered adult head model, and the partial optical path length in the brain is predicted from the proposed simplified models. The predicted partial optical path length is compared with that calculated from the five layered model to evaluate the proposed modeling.

A three-dimensional (3D) finite-element optical tomographic reconstruction algorithm based on a diffusion equation approximation is presented. The algorithm uses a regularized Newton method to update an initial (guess) optical property distribution iteratively in order to minimize an object function composed of a weighted sum of the squared difference between computed and measured data, and reconstructs the spatial distribution of the absorption and scattering coefficients of turbid media/tissues using DC data. Considering the memory requirements and computational cost for 3D reconstruction, the algorithm has been parallelized with Message-Passing Interface. We have conducted both phantom and in vivo clinical experiments to evaluate our parallelized three-dimensional reconstruction algorithm. The results show that 3D volumetric images of turbid media and in vivo tissues can be successfully reconstructed.

The reconstruction of fluorescence lifetime distributions in heterogeneous turbid media and tumor-bearing animals are experimentally demonstrated by frequency-domain measurements. A set of coupled diffusion equations are used to describe the propagation of excitation and fluorescent emission light in multiply scattering media. A finite element based reconstruction algorithm combined with Marquardt and Tikhonov regularization methods are used to obtain the fluorescence images. The experimental set-up is an automatic multi-channel frequency-domain system. 16 sources and 16 detectors are used. Experiments are performed using indocyanine green (ICG) and 3,3'-diethylthiatricarbocyanine iodide (DTTCI) in tissue-like phantoms of both single- and multi-target configurations with considerations of perfect and imperfect uptake of fluorescence dyes in the scattering media. ICG are used in tumor-bearing animal studies. Our results show that the fluorescence lifetime image of the heterogeneities within a circular surrounding medium and in-vivo tissue can be reconstructed successfully.

We have constructed a nearly real-time combined imager suitable for simultaneous ultrasound and near infrared (NIR) diffusive light imaging and co-registration. The imager consists of a combined probe and associated electronics for data acquisition. A 2D ultrasound array occupies the center of the combined probe, while 12 dual wavelength laser source fibers (780 nm and 830 nm) and 8 optical detector fibers are deployed in the periphery. We have experimentally evaluated the effects of missing optical sensors in the middle of the combined probe upon the accuracy of the reconstructed optical absorption coefficient, and assessed the improvements of reconstructed absorption coefficient with the guidance of the co-registered ultrasound. The results have shown that when the central ultrasound array area is in the neighborhood of 2 X 2 cm², which corresponds to the size of most commercial ultrasound transducers, the quality of optical images will not be degenerated. In addition to the acoustic information for cancer discrimination, NIR image reconstruction becomes much easier and more reliable. According to our results, the iterative inversion algorithm converges very fast with the guidance of a priori target temporal and spatial distributions. Only one iteration is needed to reconstruct accurate optical absorption coefficient.

Near infrared topographic imaging is effective to obtain the functional image of brain-cortex. The image is reconstructed from the change in light intensity detected with multi- channel source-detector pairs. Since the volume of brain tissue sampled by each source-detector pair is affected the scattering of tissue, it is important to evaluate the effect of scattering on the spatial resolution of the reconstructed image. In this study, the absorption change of various sizes at various positions in the adult head phantom is reconstructed by multi-channel measurements to evaluate the spatial resolution of topographic imaging. Since the heterogeneity of tissue, especially presence of low scattering CSF layer, affects the light propagation in the adult brain, the phantom consists of three layers. The diameter of the absorber is varied from 10 to 20 mm. The position of reconstructed absorber in the topographic image agrees well with that of the absorber in the brain layer of the phantom. The size of reconstructed absorber in the images is considerably broader than that of the absorber.

A multichannel time-resolved system for medical optical tomography has been developed. Its main components are an eight channel single-photon counting detector and an ultrafast picosecond solid state laser. Several instrumental aspects concerning the optical coupling between the MCP-PMT and the optical fiber bundles, the electronic components and the roto-translation stages control have been examined. Moreover the accuracy limits in the determination of optical properties of phantom objects have been largely investigated taking into account the role played by the theoretical description of light propagation, the instrumental response function (by means of a convolution procedure) and the fitting algorithm. Information on image quality parameters (spatial resolution, contrast and SNR) have been also obtained by measuring Edge Response Function in different experimental conditions. In such a work, aqueous Intralipid solutions have been used as phantom objects. Furthermore, experimental investigations on well controlled inhomogeneous with different optical properties embedded in Intralipid solutions have been carried on. As last step of the work, preliminary imaging of cylindrical objects has been performed and a Linux version of TOAST code has been used in the imaging reconstruction.

Dynamic processes in biology are often controlled by multiple parameters that interact in a complex nonlinear fashion. Increasingly, evidence has accumulated that such behavior exhibits the property of sensitivity to initial conditions, a feature exhibited by chaotic systems. One such system is the vasculature. In this report, we present what we believe is the first experimental demonstration ever of imaging chaotic behavior of the vasculature in a large tissue structure (i.e., the human forearm). Supporting these findings are results from numerical simulation demonstrating our ability to image and correctly characterize complex dynamic behavior in dense scattering media that experience spatiotemporally coincident variations in hemodynamic states.

We present preliminary results of non-invasive, near-infrared measurements of venous saturation (SvO₂) on the leg muscle of three anesthetized piglets. We have quantified the local SvO2 by analyzing the optical spectrum of the amplitude of the absorption oscillations synchronous with breathing. To induce a variation in the muscle SvO₂, we performed measurements during a protocol involving a cyclic change in the fraction of oxygen inspired by the piglet over the range 10-100% (by volume). In all three piglets, we have found a good agreement between the SvO₂ values measured non-invasively with near-infrared spectroscopy (NIRS) and those measured invasively by the analysis of venous blood samples.

Changes in phase that occurs with changes in amplitude impose severe limitations on the accuracy of frequency domain near infrared spectrometers. Phase is related to the photon pathlength in tissue and phase errors introduced by the instrument can be interpreted as changes in tissue oxygenation. The instrument described employs a reference RF modulated laser diode to eliminate the effects of amplitude- phase crosstalk and requires no feedback. Light from the reference and medium channels follow a common path through the detector and so the same phase error is imposed on both. Summing the reference and medium phase eliminates the crosstalk and enables the resultant to be attributed only to photon pathlength within the medium. It is also demonstrated that elimination of amplitude-phase crosstalk is a natural consequence of a phased array configuration.

The absorption and reduced scattering coefficients of tumours of the human breast, liver and kidney and their normal surrounding tissue have been measured in vitro for the near-infrared wavelengths between 600 and 1000 nm as well as 1064 nm. The Monte Carlo inversion technique (Simpson et al) was used to determine the optical coefficients of tissue samples from measurements of the diffuse transmittance and reflectance. The measurements of the diffusely transmitted and reflected intensities were performed using a single integrating sphere 'comparison' method. Four post-mortem samples of both liver adenocarcinoma and normal liver tissue were obtained from one subject and four samples of both tumour and normal kidney tissue were obtained from another subject. Four samples of both breast tumour and normal tissue were obtained from two patients. The scattering coefficient of tumours was found in each case to be significantly higher than that of nondiseased tissue. The absorption coefficient of tumours was generally much smaller than those of normal tissue. The scattering coefficient of tumours was 20% to 200% higher depending on the type of cancer and the wavelength, while the absorption coefficient of tumours was as much as twenty times smaller compared to normal tissue.

The objective of this paper is to study the distribution of the near infrared contrast agents delivered to mice bearing with small tumors in subsurface. Two kinds of contrast agents, NN382:ex760nm/em800nm and CY5.5:ex660nm/em700nm, which were activated by specific enzymes in tumors after bounded with certain delivery vehicle, were studied by using the 3D high spatial resolution image system. Flavoprotein and Nicotinamide adenine dinucleotide in tissue are measured simultaneously as intrinsic agents for study the effect of extrinsic contrast agents on redox state in tumor.