This PDF file contains the front matter associated with SPIE Proceedings Volume 7369, including the Title Page, Copyright information, Table of Contents, and the Conference Committee listing.

We evaluated the vascular response correlated to neural activity within a working memory "n-back" task in a population of healthy volunteers by means of time-resolved near-infrared functional spectroscopy and Generalized Linear Models. Moreover, we attempted a separation of purely cortical activation from non-cerebral contribution.

The objective of the study was to assess the usability of a near-infrared spectroscopy (NIRS) device in multimodal measurements. We combined NIRS with electroencephalography (EEG) to record hemodynamic responses and evoked potentials simultaneously, and with transcranial magnetic stimulation (TMS) to investigate hemodynamic responses to repetitive TMS (rTMS). Hemodynamic responses and visual evoked potentials (VEPs) to 3, 6, and 12 s stimuli consisting of pattern-reversing checkerboards were successfully recorded in the NIRS/EEG measurement, and ipsi- and contralateral hemodynamic responses to 0.5, 1, and 2 Hz rTMS in the NIRS/TMS measurement. In the NIRS/EEG measurements, the amplitudes of the hemodynamic responses increased from 3- to 6-s stimulus, but not from 6- to 12-s stimulus, and the VEPs showed peaks N75, P100, and N135. In the NIRS/TMS measurements, the 2-Hz stimulus produced the strongest hemodynamic responses compared to the 0.5- and 1-Hz stimuli. In two subjects oxyhemoglobin concentration decreased and in one increased as a consequence of the 2-Hz rTMS. To locate the origin of the measured NIRS responses, methods have to be developed to investigate TMS-induced scalp muscle contractions. In the future, multimodal measurements may prove useful in monitoring or treating diseases such as stroke or Alzheimer's disease.

We propose a novel method for reconstruction of fluorophore spatial distribution in a highly scattering object imaged by means of fluorescence diffuse tomography. The method is based on algebraic reconstruction principle combined with a new theoretical model of light propagation in biotissues and Monte Carlo simulations. Monte Carlo simulations are also implemented for initial data calculations in approbation of the method. The reconstruction method was tested both for simulated and experimental initial data. We demonstrate that the method provides determination of center position of fluorescent inclusion in transversal direction with the accuracy of 1 mm, in-depth direction with the accuracy of 1.5 mm and determination of size of fluorescent inclusion with the accuracy of 1.5 mm.

This paper presents detailed computational aspects of a new 3D modeling for solving the direct problem in a no-contact time-resolved Fluorescent Diffuse Optical Tomography (FDOT) method that rely on near-infrared scattered and fluorescent photons to image the optical properties and distribution of fluorescent probes in small laboratory animals. An optical scanner allowing performing in-vivo measurements in no-contact scheme was built in our laboratory and is presented. We use the three-dimensional Finite Element Method (FEM) to solve the coupled diffusion equations of excitation and fluorescence photons in highly scattering objects. The computed results allowed yielding photon density maps and the temporal profiles of photons on the surface of the small animal. Our 3D modeling of propagation of photons in the void space between the surface of the object and the detectors allows calculating the quantity of photons reaching the optodes. Simulations were carried-out on two test objects: a resin cylinder and a mouse phantom. The results demonstrate the potential applications of the method to pre-clinical imaging.

Based on a multigrid forward solver of radiative transfer equation for optical imaging, an efficient multilevel simultaneous reconstruction of absorption and scattering coefficient is presented, in which L₁ minimization can be used to localize the unknowns, especially in the presence of sparse unknowns.

Light propagation in semi-infinite and finite N-layered turbid media is studied in the steady-state, frequency, and time domains. Solutions of the diffusion equation are derived and compared to Monte Carlo simulations. In general, good agreement between the results obtained from both methods is observed.

A promising method achieving rapid convergence for image reconstruction is introduced for the continuous-wave NIR-DOT. An approach employs a constraint based on Lorentzian distributed function incorporated into Tikhonov regularization, thereby rapidly converging a stable solution.

Analytical model of radiation in highly scattering medium has been developed that describes distribution of scattered photons at arbitrary distance from the directed source. The model operates with minimum number of optical parameters of the medium, such as absorption coefficient and reduced scattering coefficient, the same as the formulas of the diffusion approximation of radiation transport equation. The validity of the model is confirmed by comparing the theoretical data with the results of Monte Carlo simulations.

The photon average trajectory method has been recently investigated as a fast reconstruction technique for time-domain diffuse optical tomography. The main disadvantage of this method is that it reconstructs the tomograms blurred due to averaging over the spatial distributions of photons. To get information about actual boundary and shape of optical inhomogeneities being reconstructed, we propose the segmentation approach based on the generation of nonlinear analytical and statistical functions of correspondence between image intensity and color space. It is shown that for simple models (absorbing macro-inhomogeneities in a homogeneous scattering medium) the proposed approach allows the true structure of inhomogeneities to be reproduced almost completely. If a medium contains randomly inhomogeneous component, our segmentation method may give artifacts which should be removed on the basis of a priori information.

Quantitative assessment of skin chromophores in a non-invasive fashion is often desirable. Especially pixel wise assessment of blood volume and blood oxygenation is beneficial for improved diagnostics. We utilized a multi-spectral imaging system for acquiring diffuse reflectance images of healthy volunteers' lower forearm. Ischemia and reactive hyperemia was introduced by occluding the upper arm with a pressure cuff for 5min with 180mmHg. Multi-spectral images were taken every 30s, before, during and after occlusion. Image reconstruction for blood volume and blood oxygenation was performed, using a two layered skin model. As the images were taken in a non-contact way, strong artifacts related to the shape (curvature) of the arms were observed, making reconstruction of optical / physiological parameters highly inaccurate. We developed a curvature correction method, which is based on extracting the curvature directly from the intensity images acquired and does not require any additional measures on the object imaged. The effectiveness of the algorithm was demonstrated, on reconstruction results of blood volume and blood oxygenation for in vivo data during occlusion of the arm. Pixel wise assessment of blood volume and blood oxygenation was made possible over the entire image area and comparison of occlusion effects between veins and surrounding skin was performed. Induced ischemia during occlusion and reactive hyperemia afterwards was observed and quantitatively assessed. Furthermore, the influence of epidermal thickness on reconstruction results was evaluated and the exact knowledge of this parameter for fully quantitative assessment was pointed out.

Temporal propagation of sinusoidally modulated light in in-homogeneous diffusive samples is investigated experimentally and by finite element simulations, showing that the amplitude and the phase of the sinusoid are affected by the presence of inclusions. Enhancement of imaging resolution with high spatial frequencies and early time-gating is shown. Detection of the phase of the modulated light is proposed as a new method for the accurate localization of tissue in-homogeneities. A fast 3-dimensional reconstruction based on the detection of spatially modulated light at a limited number of spatial frequencies is discussed.

We propose a system for 3D tomography using a single pulsed source and a time-gated camera for functional imaging studies. Reconstructions were based on a linear model based on small perturbation assumption, applying Tikhonov regularization. This approach was tested against simulations, demonstrating both detection and localization capabilities. Preliminary measurements on realistic inhomogeneous phantoms showed good detection sensibility, even for a low optical contrast, but poorer localization properties, possibly due to the still low SNR of the system. Finally, an initial in vivo test on a motor cortex activation paradigm is presented.

Within the diffusion approximation, we recently showed that the classical measurable quantity models can lead to significant deviations. Here, we show that the choice of the measurable quantity model can impact significantly the reconstructions in fluorescence diffuse optical tomography. The problem arises when i) the extrapolated boundary conditions are used and when ii) low diffusing media are considered.

A fluorescence diffuse optical tomography instrument including a dedicated reconstruction scheme which accounts for the medium optical heterogeneities is presented. It allows non-contact measurements and does not require animal immersion in an optical adaptation liquid.

We discuss the spectral distortions occurring when time-resolved diffuse spectroscopy is performed illuminating with a spectrally wide source. We show that the spectral region within the source bandwidth that exhibits the lowest absorption will dominate the resulting time-resolved curve, leading to significant distortions on the retrieved absorption spectrum (including shifts in peak positions). Due to the nonlinear behavior of the light attenuation due to absorption, this effect becomes more pronounced when including longer and longer photon path lengths. First, a theoretical treatment of the problem is given and then the distortion is described by timeresolved reflectance simulations and experimental measurements of lipid and water samples. Finally, a spectrally constrained data analysis is proposed to overcome the distortion and improve the accuracy of the estimation of chromophore concentrations from absorption spectra. Measurements on a lipid sample show a reduction of the error from 30% to 6%.

The concentration changes in oxygenated and deoxygenated haemoglobin in the exposed cortex of guinea pigs evoked by the auditory stimulation are measured by the multi-spectral imaging to investigate the relationship between spatial extent of the brain activation determined by the statistical analysis and the SNR of the concentration changes in oxygenated and deoxygenated haemoglobin. The SNR of the concentration change in oxygenated haemoglobin measured by the multi-spectral imaging is generally greater than that of deoxygenated haemoglobin. The difference in SNR tends to affect the result of the spatial extent of brain activation estimated from the changes in oxygenated and deoxygenated haemoglobin. The influence of the SNR on the spatial extent is evaluated by a numerical experiment. The results of the numerical experiment are compared with the spatial extent of the brain activation estimated from the changes in oxygenated and deoxygenated haemoglobin in the exposed cortex of guinea pigs evoked by the auditory stimulation. It is found that the spatial extent of the brain activation decreases with a decrease in SNR of the concentration change. The difference in spatial extent of the brain activation estimated from the concentration changes in oxygenated and deoxygenated haemoglobin is affected by the SNR of signal.

Angle-resolved ellipsometric data are recorded on light scattering and provide a real time process for selective imaging in scattering media. Surface and bulk effects are separated and could be used for a selective screening inside the tissues.

Statistical functions help judging optical parameters estimation (absorption μa and diffusion μs') at different experimental conditions. We show that the reduced χ² function is a powerful tool for judging goodness of fit and finding fitting interval, providing enough photons are acquired. Situations of small scattering or strong absorption, small interfiber distance, surface influence and fit boundaries influence with both infinite and semi-infinite diffusion model are analyzed together with the above statistical tools to judge the estimation. ©2009 Optical Society of America

Diffuse optical spectroscopic imaging (DOSI) is a technique to assess the spatial variation in absorption and scattering properties of the biological tissues and provides the measurement of changes in concentrations of oxy-hemoglobin and deoxy-hemoglobin. In experiments, the hemodynamics temporal evolution of vessel occlusions are observed with in vivo measurements form normal subjects and some patients in intensive care unit.

The present work will serve in a diffuse optical tomography (DOT) scanner that we are developing for small animal non-contact molecular imaging. We present a new method for deconvoluting time-domain signals for use in DOT. Time-domain signals represents reemitted light intensity as a function of time when the medium is excited by ultra-short laser pulses. Actually, each signal equals the convolution between the light propagation in the medium and the impulse response of the detection system, so-called the instrument response function (IRF). Moreover, Poisson noise present in the system has to be considered. Time-domain signals directly depend on the optical properties of a medium and so contain additional information (compared to continuous-wave signals) that should be exploited in reconstruction algorithms. As an advantage, our deconvolution method does not use a priori information about the signal. It is important to remove the IRF and noise from measured signals in order to keep only the true signal, which has a direct link to medium properties.

A 3D reconstruction of light sources in biological medium with spatial filtering and updating of the forward model is simulated numerically with use of the optical properties of mouse. The spatial filters locate the strong sources in the medium. The estimated source strength is used to remove the source positions which are not contributed to the measurement data, and improved the spatial resolution of spatial filtering. The forward model updated by spatial filtering is useful to reduce the noises by using singular value decomposition. By removing the noise space from the measurement data, the noises are reduced. The numerical simulations with a numerical phantom with the optical properties of mouse showed that the proposed method succeeded in localizing the sources in a 3D medium and that the additional noise reduction improved the reconstructed images. The proposed reconstruction scheme including the noise reduction achieves robust reconstruction of the light source distribution from the noisy measurement data and incorrect information of the optical properties of the measured subject.

Using a new algorithm of Monte-Carlo simulation, we simulate the propagation of scattered light under the restricted optical configurations which is corresponds to practical experimental setups of OCTs and DOTs. We demonstrate the path-length distributions of scattered light and the dependence of the distributions of scattering points on the path-length and the scattering order. From numerical results, we discuss the dependence of propagation of scattered light on the scattering order and path-length. This algorithm can speedy and accurately simulate the multiple scattering phenomena under the restricted optical configurations compared with the conventional Monte-Carlo simulation.

A hybrid heuristic time dependent analytical solution of the radiative transfer equation for the slab geometry is derived. Comparisons with the results of Monte Carlo simulations have shown an excellent behavior of the model in describing photon migration at short distances and early times where the solution of the diffusion equation is subjected to strong approximations.

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. Furthermore, we validated our results with preliminary measurements on calibrated diffusive solutions. 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 developed and optimized a multichannel dual-wavelength time-domain brain oximeter for functional studies in the clinical environment. The system, mounted on a 19"-rack, is interfaced with instrumentation for monitoring physiological parameters and for stimuli presentation.

The changes in cortical blood flow and blood volume of guinea pigs during auditory stimulation are measured by optical imaging systems. In this study, the change in blood flow distribution was measured by the laser speckle method and the change in blood volume was measured by the multi-spectral imaging system. The significant increase in blood flow and volume was observed around one side of the auditory area just after the onset of the stimulation. The decrease in blood volume around the other side of the auditory area was observed whereas the blood flow surrounding the auditory area is decreased during the post-resting period.

A full three-dimensional, featured-data algorithm for time-domain diffuse fluorescence tomography is presented, which inverts the Laplace-transformed time-domain coupled diffusion equations and employs a pair of real-domain transform-factors to effectively separate the fluorescent yield and lifetime parameters. By use of a multi-channel time-correlation single photon counting system and a normalized Born formulation for the inversion, the proposed scheme is experimentally validated to achieve simultaneous reconstruction of the fluorescent yield and lifetime distributions with a reasonable accuracy.

For various size, location and contrast of imitated tumors, both numerical computation and experimental validation were conducted to investigate and conclude diagnosis limitation of an NIR-DOI system.

A soft- and hardware realization of optoelectronic intellectual sensor for biomedical noninvasive studies based on the analysis of light reflected from living tissues has been described. The main feature of the developed model is use of an adaptive cross-correlation detector controlled by the digital signal processor. Algorithms and operating mode of detector are defined by the type of particular problem to be solved and conditions of measurements. The proposed model was tested to identify dynamic signals in the following areas: pulsometer, evaluation variability of the cordial rhythm, evaluation of blood saturation by oxygen.

Light propagation in an n-layered model of the brain was investigated using solutions of the diffusion theory. A neural network was applied to study the inverse problem. The reduced scattering coefficient and the absorption coefficient of the brain were determined in a four-layered model. In addition, without knowing exactly the optical properties of the other layers we calculated the absorption coefficient of the brain in a four- and five-layered model.

Importance of anatomical background model in reconstructing absorptive perturbations at different depths in the neonatal head was assessed using Monte Carlo simulations. Results suggest that prior information of the optical background can improve reconstructions, even when optical parameters are only approximately known.

We experimentally characterized the angular distribution and proportion of minimally deviated quasi-ballistic (snake) photons versus multiply scattered photons in a homogenous turbid medium. The study examined the angular distribution of photons propagating through and exiting the highly scattering medium over a narrow range about the axis of a collimated light source in trans-illumination mode. The measurements were made using an angular domain imaging system that employed one of five silicon micro-machined arrays of micro-tunnels each with a range of different acceptance angles and micro-tunnel structures. The balance between quasi-ballistic photons and unwanted multiply scattered photons accepted by the micro-machined angular filters was measured in order to determine the optimum range of acceptance angles for the system. The experiments were performed in tissue mimicking phantoms using a 2-cm thick optical cell with 0.25% Intralipid^TM and a near infrared laser. This paper also presents experimental results of the angular domain imaging system employing novel micro-tunnel arrays with minimal internal reflection which can accept the non-scattered light exiting from the turbid medium within its small acceptance angle more efficiently. Our experiments reveal that image contrast was improved from 20% to 30% by employing an angular filter array with minimal internal reflection compared to conventional square-shaped filter arrays of identical size.

Recent studies in bioluminescence tomography have been mainly focusing on the ill-posedness and non-uniqueness nature of the problem. In this paper, we demonstrate the significance of the effect of bioluminescence source decay on the reconstruction results with both simulations and phantom studies. Time variation of the bioluminescence source can cause artifacts in the tomographic images. The results show that source dynamic information is pivotal for accurate reconstruction when the decay half-life is comparable to the duration of the data acquisition.

Using extracted spectral features is proposed to reconstruct video-rate optical-properties images. Compared with reconstruction through time-sequence data, the results through spectral features are exempt from noise affection, and are able to differentiate hemodynamic conditions in a single heart-beat cycle.