Accuracy limits in the determination of absolute optical properties using time-resolved NIR spectroscopy

Fluorescence diffuse optical monitoring of bioreactors: a hybrid deep learning and model-based approach for tomography

Biosynthesis in bioreactors plays a vital role in many applications, but tools for accurate in situ monitoring of the cells are still lacking. By engineering the cells such that their conditions are reported through fluorescence, it is possible to fill in the gap using fluorescence diffuse optical tomography (fDOT). However, the spatial accuracy of the reconstruction can still be limited, due to e.g. undersampling and inaccurate estimation of the optical properties. Utilizing controlled phantom studies, we use a two-step hybrid approach, where a preliminary fDOT result is first obtained using the classic model-based optimization, and then enhanced using a neural network. We show in this paper using both simulated and phantom experiments that the proposed method can lead to a 8-fold improvement (Intersection over Union) of fluorescence inclusion reconstruction in noisy conditions, at the same speed of conventional neural network-based methods. This is an important step towards our ultimate goal of fDOT monitoring of bioreactors.

Two-layered blood-lipid phantom and method to determine absorption and oxygenation employing changes in moments of DTOFs

Near-infrared spectroscopy (NIRS) is an established technique for measuring tissue oxygen saturation (StO₂), which is of high clinical value. For tissues that have layered structures, it is challenging but clinically relevant to obtain StO₂ of the different layers, e.g. brain and scalp. For this aim, we present a new method of data analysis for time-domain NIRS (TD-NIRS) and a new two-layered blood-lipid phantom. The new analysis method enables accurate determination of even large changes of the absorption coefficient (Δµ_a) in multiple layers. By adding Δµ_a to the baseline µ_a, this method provides absolute µ_a and hence StO₂ in multiple layers. The method utilizes (i) changes in statistical moments of the distributions of times of flight of photons (DTOFs), (ii) an analytical solution of the diffusion equation for an N-layered medium, (iii) and the Levenberg-Marquardt algorithm (LMA) to determine Δµ_a in multiple layers from the changes in moments. The method is suitable for NIRS tissue oximetry (relying on µ_a) as well as functional NIRS (fNIRS) applications (relying on Δµ_a). Experiments were conducted on a new phantom, which enabled us to simulate dynamic StO₂ changes in two layers for the first time. Two separate compartments, which mimic superficial and deep layers, hold blood-lipid mixtures that can be deoxygenated (using yeast) and oxygenated (by bubbling oxygen) independently. Simultaneous NIRS measurements can be performed on the two-layered medium (variable superficial layer thickness, L), the deep (homogeneous), and/or the superficial (homogeneous). In two experiments involving ink, we increased the nominal µ_a in one of two compartments from 0.05 to 0.25 cm^-1, L set to 14.5 mm. In three experiments involving blood (L set to 12, 15, or 17 mm), we used a protocol consisting of six deoxygenation cycles. A state-of-the-art multi-wavelength TD-NIRS system measured simultaneously on the two-layered medium, as well as on the deep compartment for a reference. The new method accurately determined µ_a (and hence StO₂) in both compartments. The method is a significant progress in overcoming the contamination from the superficial layer, which is beneficial for NIRS and fNIRS applications, and may improve the determination of StO₂ in the brain from measurements on the head. The advanced phantom may assist in the ongoing effort towards more realistic standardized performance tests in NIRS tissue oximetry. Data and MATLAB codes used in this study were made publicly available.

Reconstruction of optical coefficients in turbid media using time-resolved reflectance and calibration-free instrument response functions

Measurements of time-resolved reflectance from a homogenous turbid medium can be employed to retrieve the absolute values of its optical transport coefficients. However, the uncertainty in the temporal shift of the experimentally determined instrument response function (IRF) with respect to the real system response can lead to errors in optical property reconstructions. Instrument noise and measurement of the IRF in a reflectance geometry can exacerbate these errors. Here, we examine three reconstruction approaches that avoid requiring direct measurements of photon launch times. They work by (a) fitting relative shapes of the reflectance profile with a pre-determined constraint on the scattering coefficient, (b) calibrating launch-time differences via a reference sample, and (c) freely fitting for the launch-time difference within the inverse problem. Analysis methods that can place a tight bound on the scattering coefficient can produce errors within 5-15% for both absorption and scattering at source-detector separations of 10 and 15 mm. Including the time-shift in the fitting procedure also recovered optical coefficients to under 20% but showed large crosstalk between extracted scattering and absorption coefficients. We find that the uncertainty in the temporal shift greatly impacts the reconstructed reduced scattering coefficient compared to absorption.

Time-domain NIRS system based on supercontinuum light source and multi-wavelength detection: validation for tissue oxygenation studies

We present and validate a multi-wavelength time-domain near-infrared spectroscopy (TD-NIRS) system that avoids switching wavelengths and instead exploits the full capability of a supercontinuum light source by emitting and acquiring signals for the whole chosen range of wavelengths. The system was designed for muscle and brain oxygenation monitoring in a clinical environment. A pulsed supercontinuum laser emits broadband light and each of two detection modules acquires the distributions of times of flight of photons (DTOFs) for 16 spectral channels (used width 12.5 nm / channel), providing a total of 32 DTOFs at up to 3 Hz. Two emitting fibers and two detection fiber bundles allow simultaneous measurements at two positions on the tissue or at two source-detector separations. Three established protocols (BIP, MEDPHOT, and nEUROPt) were used to quantitatively assess the system's performance, including linearity, coupling, accuracy, and depth sensitivity. Measurements were performed on 32 homogeneous phantoms and two inhomogeneous phantoms (solid and liquid). Furthermore, measurements on two blood-lipid phantoms with a varied amount of blood and Intralipid provide the strongest validation for accurate tissue oximetry. The retrieved hemoglobin concentrations and oxygen saturation match well with the reference values that were obtained using a commercially available NIRS system (OxiplexTS) and a blood gas analyzer (ABL90 FLEX), except a discrepancy occurs for the lowest amount of Intralipid. In-vivo measurements on the forearm of three healthy volunteers during arterial (250 mmHg) and venous (60 mmHg) cuff occlusions provide an example of tissue monitoring during the expected hemodynamic changes that follow previously well-described physiologies. All results, including quantitative parameters, can be compared to other systems that report similar tests. Overall, the presented TD-NIRS system has an exemplary performance evaluated with state-of-the-art performance assessment methods.

Efficient numerical modelling of time-domain light propagation in curved 3D absorbing and scattering media with finite differences

An efficient approach is introduced for modelling light propagation in the time domain in 3D heterogeneous absorbing and scattering media (e.g. biological tissues) with curved boundaries. It relies on the finite difference method (FDM) in conjuction with the Crank-Nicolson method for accurately solving the optical diffusion equation (DE). The strength of the FDM lies in its simplicity and efficiency, since the equations are easy to set up, and accessing neighboring grid points only requires simple memory operations, leading to efficient code execution. Owing to its use of Cartesian grids, the FDM is generally thought cumbersome compared to the finite element method (FEM) for dealing with media with curved boundaries. However, to apply the FDM to such media, the blocking-off method can be resorted to. To account for the change of the refractive index at the boundary, Robin-type boundary conditions are considered. This requires the computation of surface normals. We resort here for the first time to the Sobel operator borrowed from image processing to perform this task. The Sobel operator is easy to implement, fast, and allows obtaining a smooth field of normal vectors along the boundary. The main contribution of this work is to arrive at a complete numerical FDM-based model of light propagation in the time domain in 3D absorbing and scattering media with curved geometries, taking into account realistic refractive index mismatch boundary conditions. The fluence rate obtained with this numerical model is shown to reproduce well that obtained with independent gold-standard Monte Carlo simulations.

Effect of adipose tissue thickness and tissue optical properties on the differential pathlength factor estimation for NIRS studies on human skeletal muscle

We propose a quantitative and systematic investigation of the differential pathlength factor (DPF) behavior for skeletal muscles and its dependence on different factors, such as the subcutaneous adipose tissue thickness (ATT), the variations of the tissue absorption (µ_a ) and reduced scattering (µ'_s ) coefficients, and the source-detector distance. A time domain (TD) NIRS simulation study is performed in a two-layer geometry mimicking a human skeletal muscle with an overlying adipose tissue layer. The DPF decreases when µ_a increases, while it increases when µ'_s increases. Moreover, a positive correlation between DPF and ATT is found. These results are supported by an in-vivo TD NIRS study on vastus lateralis and biceps brachii muscles of eleven subjects at rest, showing a high inter-subject and inter-muscle variability.

Direct estimation of the reduced scattering coefficient from experimentally measured time-resolved reflectance via Monte Carlo based lookup tables

A heuristic method for estimating the reduced scattering coefficient (µ_s') of turbid media using time-resolved reflectance is presented. The technique requires measurements of the distributions of times-of-flight (DTOF) of photons arriving at two identical detection channels placed at unique distances relative to a source. Measured temporal shifts in DTOF peak intensities at the two channels were used to estimate µ_s' of the medium using Monte Carlo (MC) simulation-based lookup tables. MC simulations were used to compute temporal shifts in modeled reflectance at experimentally employed source-detector separations (SDS) for media spanning a wide range of optical properties to construct look up tables. Experiments in Intralipid (IL) phantoms demonstrated that we could retrieve µ_s' with errors ranging between 6-25% of expected (literature) values, using reflectance measured across 650-800 nm and SDS of 5-15 mm. Advantages of the technique include direct processing of measured data without requiring iterative non-linear curve fitting. We also discuss applicability of this approach for media with low scattering coefficients where the commonly employed diffusion theory analysis could be inaccurate, with practical recommendations for use.

Instrument response function acquisition in reflectance geometry for time-resolved diffuse optical measurements

In time-domain diffuse optical spectroscopy, the simultaneous acquisition of the time-of-flight distribution (DTOF) of photons traveling in a diffusive medium and of the instrument response function (IRF) is necessary to perform quantitative measurements of optical properties (absorption and reduced scattering coefficients) while taking into account the non-idealities of a real system (e.g. temporal resolution and time delays). The IRF acquisition can be a non-trivial and time-consuming operation that requires directly facing the injection and collection fibers. Since this operation is not always possible, a new IRF measurement scheme is here proposed where the IRF is acquired in reflectance geometry from a corrugate reflective surface. Validation measurements on a set of reference homogenous phantoms have been performed, resulting in an error in the optical properties estimation lower than 10% with respect to the typical IRF configuration. Thus, the proposed method proved to be a reliable approach that after a preliminary calibration can be exploited in a laboratory and clinical set-ups, leading to faster and more accurate measurements and reducing the operator-dependent performance.

Method to improve the depth sensitivity of diffuse reflectance measurements to absorption changes in optically turbid medium

We have studied the spatial distributions of the sensitivity of time-resolved near-infrared diffuse reflectance measurement. Sensitivity factors representing a change of parameters of a measured optical signal induced by absorption perturbation in a certain voxel of the medium were simulated using the diffusion equation solution. The parameters were statistical moments of measured distributions of time of flight of photons (DTOFs) i.e., the total number of photons, mean time of flight and variance. The distributions of the sensitivity of statistical moments of DTOFs to a change in absorption were generated for various source-detector separations and various optical properties of the medium. Furthermore, differential sensitivity distributions for two different source-detector separations were calculated. A measurement geometry, in which two detection spots, separated by 5 mm, in combination with two sources was proposed. For this setup differences between the signals obtained for both detectors were calculated independently for both sources and afterward summed up for both source positions. Obtained differences in moments of DTOFs assessed at two source-detector separations and summed up for different positioning of the sources allowed to shape up the sensitivity profiles. Calculated sensitivity profiles show that positive sensitivities of the mean time of flight of photons and variance of the DTOF can be obtained. These positive sensitivity areas are located just between both detection spots and cover the compartment located deeply in the medium. The sensitivity in superficial compartments of the medium is negative and much smaller in amplitude. The proposed technique can be used for improved discrimination of optical signals related to the intracerebral change in absorption which remains a serious obstacle in the application of the NIRS technique in the assessment of brain oxygenation or perfusion.

Head model based on the shape of the subject's head for optical brain imaging

Optical imaging methods such as near-infrared spectroscopy and diffuse optical tomography rely on models to solve the inverse problem. Imaging an adult human head also requires a head model. Using a model, which makes describing the structure of the head better, leads to acquiring a more accurate absorption map. Here, by combining the key features of layered slab models and head atlases, we introduce a new two-layered head model that is based on the surface geometry of the subject's head with variable thickness of the superficial layer. Using the Monte Carlo approach, we assess the performance of our model for fitting the optical properties from simulated time-resolved data of the adult head in a null distance source-detector configuration. Using our model, we observed improved results at 70 percent of the locations on the head and an overall 20 percent reduction in relative error compared to layered slab model.

Accuracy and precision of tissue optical properties and hemodynamic parameters estimated by the BabyLux device: a hybrid time-resolved near-infrared and diffuse correlation spectroscopy neuro-monitor

We have investigated the accuracy and precision of "the BabyLux device", a hybrid time-resolved near-infrared (TRS) and diffuse correlation spectroscopy (DCS) neuro-monitor for the pre-term infant. Numerical data with realistic noise were simulated and analyzed using the BabyLux device as a reference system and different experimental and analysis parameters. The results describe the limits for the precision and the accuracy to be expected. The dependence of these limits on different experimental conditions and choices of the analysis method is also described. Experiments demonstrate comparable values for precision with respect to the simulation results.

BabyLux device: a diffuse optical system integrating diffuse correlation spectroscopy and time-resolved near-infrared spectroscopy for the neuromonitoring of the premature newborn brain

The BabyLux device is a hybrid diffuse optical neuromonitor that has been developed and built to be employed in neonatal intensive care unit for the noninvasive, cot-side monitoring of microvascular cerebral blood flow and blood oxygenation. It integrates time-resolved near-infrared and diffuse correlation spectroscopies in a user-friendly device as a prototype for a future medical grade device. We present a thorough characterization of the device performance using test measurements in laboratory settings. Tests on solid phantoms report an accuracy of optical property estimation of about 10%, which is expected when using the photon diffusion equation as the model. The measurement of the optical and dynamic properties is stable during several hours of measurements within 3% of the average value. In addition, these measurements are repeatable between different days of measurement, showing a maximal variation of 5% in the optical properties and 8% for the particle diffusion coefficient on a liquid phantom. The variability over test/retest evaluation is < 3 % . The integration of the two modalities is robust and without any cross talk between the two. We also perform in vivo measurements on the adult forearm during arterial cuff occlusion to show that the device can measure a wide range of tissue hemodynamic parameters. We suggest that this platform can form the basis of the next-generation neonatal neuromonitors to be developed for extensive, multicenter clinical testing.

Characterization of a Time-Resolved Diffuse Optical Spectroscopy Prototype Using Low-Cost, Compact Single Photon Avalanche Detectors for Tissue Optics Applications

Time-resolved diffuse optical spectroscopy (TR-DOS) is an increasingly used method to determine the optical properties of diffusive media, particularly for medical applications including functional brain, breast and muscle measurements. For medical imaging applications, important features of new generation TR-DOS systems are low-cost, small size and efficient inverse modeling. To address the issues of low-cost, compact size and high integration capabilities, we have developed free-running (FR) single-photon avalanche diodes (SPADs) using 130 nm silicon complementary metal-oxide-semiconductor (CMOS) technology and used it in a TR-DOS prototype. This prototype was validated using assessments from two known protocols for evaluating TR-DOS systems for tissue optics applications. Following the basic instrumental performance protocol, our prototype had sub-nanosecond total instrument response function and low differential non-linearity of a few percent. Also, using light with optical power lower than the maximum permissible exposure for human skin, this prototype can acquire raw data in reflectance geometry for phantoms with optical properties similar to human tissues. Following the MEDPHOT protocol, the absolute values of the optical properties for several homogeneous phantoms were retrieved with good accuracy and linearity using a best-fitting model based on the Levenberg-Marquardt method. Overall, the results of this study show that our silicon CMOS-based SPAD detectors can be used to build a multichannel TR-DOS prototype. Also, real-time functional monitoring of human tissue such as muscles, breasts and newborn heads will be possible by integrating this detector with a time-to-digital converter (TDC).

Lawrence K. Assessment of a multi-layered diffuse correlation spectroscopy method for monitoring cerebral blood flow in adults

Diffuse correlation spectroscopy (DCS) is a promising technique for brain monitoring as it can provide a continuous signal that is directly related to cerebral blood flow (CBF); however, signal contamination from extracerebral tissue can cause flow underestimations. The goal of this study was to investigate whether a multi-layered (ML) model that accounts for light propagation through the different tissue layers could successfully separate scalp and brain flow when applied to DCS data acquired at multiple source-detector distances. The method was first validated with phantom experiments. Next, experiments were conducted in a pig model of the adult head with a mean extracerebral tissue thickness of 9.8 ± 0.4 mm. Reductions in CBF were measured by ML DCS and computed tomography perfusion for validation; excellent agreement was observed by a mean difference of 1.2 ± 4.6% (CI_95%: -31.1 and 28.6) between the two modalities, which was not significantly different.

Time-resolved subtraction method for measuring optical properties of turbid media

Near-infrared spectroscopy is a noninvasive optical method used primarily to monitor tissue oxygenation due to the absorption properties of hemoglobin. Accurate estimation of hemoglobin concentrations and other light absorbers requires techniques that can separate the effect of absorption from the much greater effect of light scattering. One of the most advanced methods is time-resolved near-infrared spectroscopy (TR-NIRS), which measures the absorption and scattering coefficients of a turbid medium by modeling the recorded distribution time of flight of photons. A challenge with TR-NIRS is that it requires accurate characterization of the dispersion caused by the system. In this study, we present a method for circumventing this problem by applying statistical moment analysis to two time-of-flight distributions measured at separated source-detector distances. Simulations based on analytical models and Monte Carlo code, and tissue-mimicking phantoms, were used to demonstrate its accuracy for source-detector distances typically used in neuroimaging applications. The simplicity of the approach is well suited to real-time applications requiring accurate quantification of the optical properties of a turbid medium.

Lawrence K. Assessment of the best flow model to characterize diffuse correlation spectroscopy data acquired directly on the brain

Diffuse correlation spectroscopy (DCS) is a non-invasive optical technique capable of monitoring tissue perfusion. The normalized temporal intensity autocorrelation function generated by DCS is typically characterized by assuming that the movement of erythrocytes can be modeled as a Brownian diffusion-like process instead of by the expected random flow model. Recently, a hybrid model, referred to as the hydrodynamic diffusion model, was proposed, which combines the random and Brownian flow models. The purpose of this study was to investigate the best model to describe autocorrelation functions acquired directly on the brain in order to avoid confounding effects of extracerebral tissues. Data were acquired from 11 pigs during normocapnia and hypocapnia, and flow changes were verified by computed tomography perfusion (CTP). The hydrodynamic diffusion model was found to provide the best fit to the autocorrelation functions; however, no significant difference for relative flow changes measured by the Brownian and hydrodynamic diffusion models was observed.

Sensitivity analysis for oblique incidence reflectometry using Monte Carlo simulations

Oblique incidence reflectometry has developed into an effective, noncontact, and noninvasive measurement technology for the quantification of both the reduced scattering and absorption coefficients of a sample. The optical properties are deduced by analyzing only the shape of the reflectance profiles. This article presents a sensitivity analysis of the technique in turbid media. Monte Carlo simulations are used to investigate the technique and its potential to distinguish the small changes between different levels of scattering. We present various regions of the dynamic range of optical properties in which system demands vary to be able to detect subtle changes in the structure of the medium, translated as measured optical properties. Effects of variation in anisotropy are discussed and results presented. Finally, experimental data of milk products with different fat content are considered as examples for comparison.

Lawrence K, Lee TY. Coupling of cerebral blood flow and oxygen consumption during hypothermia in newborn piglets as measured by time-resolved near-infrared spectroscopy: a pilot study

Hypothermia (HT) is a potent neuroprotective therapy that is now widely used in following neurological emergencies, such as neonatal asphyxia. An important mechanism of HT-induced neuroprotection is attributed to the associated reduction in the cerebral metabolic rate of oxygen ([Formula: see text]). Since cerebral circulation and metabolism are tightly regulated, reduction in [Formula: see text] typically results in decreased cerebral blood flow (CBF); it is only under oxidative stress, e.g., hypoxia-ischemia, that oxygen extraction fraction (OEF) deviates from its basal value, which can lead to cerebral dysfunction. As such, it is critical to measure these key physiological parameters during therapeutic HT. This report investigates a noninvasive method of measuring the coupling of [Formula: see text] and CBF under HT and different anesthetic combinations of propofol/nitrous-oxide ([Formula: see text]) that may be used in clinical practice. Both CBF and [Formula: see text] decreased with decreasing temperature, but the OEF remained unchanged, which indicates a tight coupling of flow and metabolism under different anesthetics and over the mild HT temperature range (38°C to 33°C).

Study of the effect introduced by an integrating sphere on the temporal profile characterization of short laser pulses propagating through a turbid medium

When a nanosecond laser pulse is transmitted through a highly scattering material, its irradiance decreases as it propagates; this is because of the spatial and temporal pulse profile stretching owing to multiple scattering events. Although the effect of temporal distortion is much less significant than that of the spatial distortion for applications where the laser beam is focused on a subsurface target (writing of waveguides, for example), it becomes significant for applications where the laser pulse must attain certain temporal width after the beam propagated is collimated through a turbid medium (photoacoustic tomography, for example). The objective of this work is to determine the transfer function associated to an integrating sphere measurement of the temporal intensity profile involving turbid media samples. The transfer function is found to be related to the geometrical characteristics of the integrating sphere and the optical properties of the turbid media. This procedure opens a new possibility for optical property characterization and enables the use of an integrating sphere for time-dependent intensity measurements.

Comparison of a layered slab and an atlas head model for Monte Carlo fitting of time-domain near-infrared spectroscopy data of the adult head

Near-infrared spectroscopy (NIRS) estimations of the adult brain baseline optical properties based on a homogeneous model of the head are known to introduce significant contamination from extracerebral layers. More complex models have been proposed and occasionally applied to in vivo data, but their performances have never been characterized on realistic head structures. Here we implement a flexible fitting routine of time-domain NIRS data using graphics processing unit based Monte Carlo simulations. We compare the results for two different geometries: a two-layer slab with variable thickness of the first layer and a template atlas head registered to the subject's head surface. We characterize the performance of the Monte Carlo approaches for fitting the optical properties from simulated time-resolved data of the adult head. We show that both geometries provide better results than the commonly used homogeneous model, and we quantify the improvement in terms of accuracy, linearity, and cross-talk from extracerebral layers.

Broadband photon time-of-flight spectroscopy of pharmaceuticals and highly scattering plastics in the VIS and close NIR spectral ranges

We present extended spectroscopic analysis of pharmaceutical tablets in the close near infrared spectral range performed using broadband photon time-of-flight (PTOF) absorption and scattering spectra measurements. We show that the absorption spectra can be used to perform evaluation of the chemical composition of pharmaceutical tablets without need for chemo-metric calibration. The spectroscopic analysis was performed using an advanced PTOF spectrometer operating in the 650 to 1400 nm spectral range. By employing temporal stabilization of the system we achieve the high precision of 0.5% required to evaluate the concentration of tablet ingredients. In order to further illustrate the performance of the system, we present the first ever reported broadband evaluation of absorption and scattering spectra from pure and doped Spectralon®.

Lawrence K. Improving the depth sensitivity of time-resolved measurements by extracting the distribution of times-of-flight

Time-resolved (TR) techniques provide a means of discriminating photons based on their time-of-flight. Since early arriving photons have a lower probability of probing deeper tissue than photons with long time-of-flight, time-windowing has been suggested as a method for improving depth sensitivity. However, TR measurements also contain instrument contributions (instrument-response-function, IRF), which cause temporal broadening of the measured temporal point-spread function (TPSF) compared to the true distribution of times-of-flight (DTOF). The purpose of this study was to investigate the influence of the IRF on the depth sensitivity of TR measurements. TPSFs were acquired on homogeneous and two-layer tissue-mimicking phantoms with varying optical properties. The measured IRF and TPSFs were deconvolved using a stable algorithm to recover the DTOFs. The microscopic Beer-Lambert law was applied to the TPSFs and DTOFs to obtain depth-resolved absorption changes. In contrast to the DTOF, the latest part of the TPSF was not the most sensitive to absorption changes in the lower layer, which was confirmed by computer simulations. The improved depth sensitivity of the DTOF was illustrated in a pig model of the adult human head. Specifically, it was shown that dynamic absorption changes obtained from the late part of the DTOFs recovered from TPSFs acquired by probes positioned on the scalp were similar to absorption changes measured directly on the brain. These results collectively demonstrate that this method improves the depth sensitivity of TR measurements by removing the effects of the IRF.

Exploiting breakdown of the similarity relation for diffuse light transport: simultaneous retrieval of scattering anisotropy and diffusion constant

As manifested in the similarity relation of diffuse light transport, it is difficult to assess single scattering characteristics from multiply scattered light. We take advantage of the limited validity of the diffusion approximation of light transport and demonstrate, experimentally and numerically, that even deep into the multiple scattering regime, time-resolved detection of transmitted light allows simultaneous assessment of both single scattering anisotropy and scattering mean free path, and therefore also macroscopic parameters like the diffusion constant and the transport mean free path. This is achieved via careful assessment of early light and matching against Monte Carlo simulations of radiative transfer.

The effect of temporal impulse response on experimental reduction of photon scatter in time-resolved diffuse optical tomography

New fast detector technology has driven significant renewed interest in time-resolved measurement of early photons in improving imaging resolution in diffuse optical tomography and fluorescence mediated tomography in recent years. In practice, selection of early photons results in significantly narrower instrument photon density sensitivity functions (PDSFs) than the continuous wave case, resulting in a better conditioned reconstruction problem. In this work, we studied the quantitative impact of the instrument temporal impulse response function (TIRF) on experimental PDSFs in tissue mimicking optical phantoms. We used a multimode fiber dispersion method to vary the system TIRF over a range of representative literature values. Substantial disagreement in PDSF width--by up to 40%--was observed between experimental measurements and Monte Carlo (MC) models of photon propagation over the range of TIRFs studied. On average, PDSFs were broadened by about 0.3 mm at the center plane of the 2 cm wide imaging chamber per 100 ps of the instrument TIRF at early times. Further, this broadening was comparable on both the source and detector sides. Results were confirmed by convolution of instrument TIRFs with MC simulations. These data also underscore the importance of correcting imaging PDSFs for the instrument TIRF when performing tomographic image reconstruction to ensure accurate data-model agreement.

Depth resolution and multiexponential lifetime analyses of reflectance-based time-domain fluorescence data

Time-domain fluorescence imaging is a powerful new technique that adds a rich amount of information to conventional fluorescence imaging. Specifically, time-domain fluorescence can be used to remove autofluorescence from signals, resolve multiple fluorophore concentrations, provide information about tissue microenvironments, and, for reflectance-based imaging systems, resolve inclusion depth. The present study provides the theory behind an improved method of analyzing reflectance-based time-domain data that is capable of accurately recovering mixed concentration ratios of multiple fluorescent agents while also recovering the depth of the inclusion. The utility of the approach was demonstrated in a number of simulations and in tissuelike phantom experiments using a short source-detector separation system. The major findings of this study were (1) both depth of an inclusion and accurate ratios of two-fluorophore concentrations can be recovered accurately up to depths of approximately 1 cm with only the optical properties of the medium as prior knowledge, (2) resolving the depth and accounting for the dispersion effects on fluorescent lifetimes is crucial to the accuracy of recovered ratios, and (3) ratios of three-fluorophore concentrations can be resolved at depth but only if the lifetimes of the three fluorophores are used as prior knowledge. By accurately resolving the concentration ratios of two to three fluorophores, it may be possible to remove autofluorescence or carry out quantitative techniques, such as reference tracer kinetic modeling or ratiometric approaches, to determine receptor binding or microenvironment parameters in point-based time-domain fluorescence applications.

Molecular imaging using light-absorbing imaging agents and a clinical optical breast imaging system--a phantom study

PURPOSE

The aim of the study was to determine the feasibility of using a clinical optical breast scanner with molecular imaging strategies based on modulating light transmission.

PROCEDURES

Different concentrations of single-walled carbon nanotubes (SWNT; 0.8-20.0 nM) and black hole quencher-3 (BHQ-3; 2.0-32.0 µM) were studied in specifically designed phantoms (200-1,570 mm(3)) with a clinical optical breast scanner using four wavelengths. Each phantom was placed in the scanner tank filled with optical matching medium. Background scans were compared to absorption scans, and reproducibility was assessed.

RESULTS

All SWNT phantoms were detected at four wavelengths, with best results at 684 nm. Higher concentrations (≥8.0 µM) were needed for BHQ-3 detection, with the largest contrast at 684 nm. The optical absorption signal was dependent on phantom size and concentration. Reproducibility was excellent (intraclass correlation 0.93-0.98).

CONCLUSION

Nanomolar concentrations of SWNT and micromolar concentrations of BHQ-3 in phantoms were reproducibly detected, showing the potential of light absorbers, with appropriate targeting ligands, as molecular imaging agents for clinical optical breast imaging.

Light propagation from fluorescent probes in biological tissues by coupled time-dependent parabolic simplified spherical harmonics equations

We introduce a system of coupled time-dependent parabolic simplified spherical harmonic equations to model the propagation of both excitation and fluorescence light in biological tissues. We resort to a finite element approach to obtain the time-dependent profile of the excitation and the fluorescence light fields in the medium. We present results for cases involving two geometries in three-dimensions: a homogeneous cylinder with an embedded fluorescent inclusion and a realistically-shaped rodent with an embedded inclusion alike an organ filled with a fluorescent probe. For the cylindrical geometry, we show the differences in the time-dependent fluorescence response for a point-like, a spherical, and a spherically Gaussian distributed fluorescent inclusion. From our results, we conclude that the model is able to describe the time-dependent excitation and fluorescent light transfer in small geometries with high absorption coefficients and in nondiffusive domains, as may be found in small animal diffuse optical tomography (DOT) and fluorescence DOT imaging.

Assessment of the frequency-domain multi-distance method to evaluate the brain optical properties: Monte Carlo simulations from neonate to adult

The near infrared spectroscopy (NIRS) frequency-domain multi-distance (FD-MD) method allows for the estimation of optical properties in biological tissue using the phase and intensity of radiofrequency modulated light at different source-detector separations. In this study, we evaluated the accuracy of this method to retrieve the absorption coefficient of the brain at different ages. Synthetic measurements were generated with Monte Carlo simulations in magnetic resonance imaging (MRI)-based heterogeneous head models for four ages: newborn, 6 and 12 month old infants, and adult. For each age, we determined the optimal set of source-detector separations and estimated the corresponding errors. Errors arise from different origins: methodological (FD-MD) and anatomical (curvature, head size and contamination by extra-cerebral tissues). We found that the brain optical absorption could be retrieved with an error between 8-24% in neonates and infants, while the error increased to 19-44% in adults over all source-detector distances. The dominant contribution to the error was found to be the head curvature in neonates and infants, and the extra-cerebral tissues in adults.

Comparison of time-resolved and continuous-wave near-infrared techniques for measuring cerebral blood flow in piglets

A primary focus of neurointensive care is monitoring the injured brain to detect harmful events that can impair cerebral blood flow (CBF), resulting in further injury. Since current noninvasive methods used in the clinic can only assess blood flow indirectly, the goal of this research is to develop an optical technique for measuring absolute CBF. A time-resolved near-infrared (TR-NIR) apparatus is built and CBF is determined by a bolus-tracking method using indocyanine green as an intravascular flow tracer. As a first step in the validation of this technique, CBF is measured in newborn piglets to avoid signal contamination from extracerebral tissue. Measurements are acquired under three conditions: normocapnia, hypercapnia, and following carotid occlusion. For comparison, CBF is concurrently measured by a previously developed continuous-wave NIR method. A strong correlation between CBF measurements from the two techniques is revealed with a slope of 0.79±0.06, an intercept of -2.2±2.5 ml∕100 g∕min, and an R2 of 0.810±0.088. Results demonstrate that TR-NIR can measure CBF with reasonable accuracy and is sensitive to flow changes. The discrepancy between the two methods at higher CBF could be caused by differences in depth sensitivities between continuous-wave and time-resolved measurements.

Development of an optical imaging platform for functional imaging of small animals using wide-field excitation

The design and characterization of a time-resolved functional imager using a wide-field excitation scheme for small animal imaging is described. The optimal operation parameters are established based on phantom studies. The performance of the platform for functional imaging and the simultaneous 3D reconstruction of absorption and scattering coefficients is investigated in vitro.

Diffuse light propagation in biological media by a time-domain parabolic simplified spherical harmonics approximation with ray-divergence effects

We present a simplified spherical harmonics approximation for the time-domain radiative transfer equation including the source-divergence effect. This leads to a set of coupled partial differential equations (PDEs) of the parabolic type that model diffuse light propagation in biological-tissue-like media. We introduce a finite element approach for solving these PDEs, thereby obtaining the time-dependent spatial profile of the fluence. We compare the results with the diffusion equation and Monte Carlo simulations. The fluence obtained via our model is shown to reproduce well the Monte Carlo results in all cases and improves on the solution of the diffusion equation in homogeneous diffusive-defying media. Our solution also shows more sensitivity than the diffusion equation to changes in the absorption coefficient of small inclusions.

Early-photon fluorescence tomography: spatial resolution improvements and noise stability considerations

In vivo tissue imaging using near-infrared light suffers from low spatial resolution and poor contrast recovery because of highly scattered photon transport. For diffuse optical tomography (DOT) and fluorescence molecular tomography (FMT), the resolution is limited to about 5-10% of the diameter of the tissue being imaged, which puts it in the range of performance seen in nuclear medicine. This paper introduces the mathematical formalism explaining why the resolution of FMT can be significantly improved when using instruments acquiring fast time-domain optical signals. This is achieved through singular-value analysis of the time-gated inverse problem based on weakly diffused photons. Simulations relevant to mouse imaging are presented showing that, in stark contrast to steady-state imaging, early time-gated intensities (within 200 ps or 400 ps) can in principle be used to resolve small fluorescent targets (radii from 1.5 to 2.5 mm) separated by less than 1.5 mm.

Intraoperative monitoring of depth-dependent hemoglobin concentration changes during carotid endarterectomy by time-resolved spectroscopy

By measuring the adult human head during carotid endarterectomy, we investigate the depth sensitivity of two methods for deriving the absorption coefficient changes (Dmu(a)) from time-resolved reflectance data to absorption changes in inhomogeneous media: (1) the curve-fitting method based on the diffusion equation (DE-fit method) and (2) the time-independent calculation based on the modified Lambert-Beer law (MLB method). Remarkable differences in the determined values of Dmu(a) caused by clamping the external carotid artery and subsequently clamping the common carotid artery were observed between the methods. The DE-fit method was more sensitive to mu(a) changes in cerebral tissues, whereas the MLB method was rather sensitive to mu(a) changes in the extracerebral tissues. Our results indicated that the DE-fit was useful for monitoring the cerebral blood circulation and oxygenation during neurosurgical operations. In addition, the combined evaluation of mu(a) changes with the DE-fit and MLB methods will provide us with more available information about the hemodynamic changes in the depth direction.

Fully automated time domain spectrometer for the absorption and scattering characterization of diffusive media

We describe a system for absorption and scattering spectroscopy of diffusive media based on time-resolved reflectance and transmittance measurements. The system is operated with mode-locked lasers tunable in the 550-1050 nm spectral range and on a detection chain based on time-correlated single-photon counting. All measurement procedures such as laser tuning and optimization, signal conditioning, data acquisition, and analysis are completely automated, permitting spectral measurements over the whole range in a few minutes. The criticalities of the system are discussed together with the strategies to compensate them. The Medphot protocol devised for the characterization of photon migration instruments was applied to assess the system performances in terms of accuracy, linearity, noise, stability, and reproducibility. Finally, an example of application of the instrument to the spectroscopy of powders is presented.

Hyperspectral diffuse reflectance imaging for rapid, noncontact measurement of the optical properties of turbid materials

We present a method and technique of using hyperspectral diffuse reflectance for rapid determination of the optical properties of turbid media. A hyperspectral imaging system in line scanning mode was used to acquire spatial diffuse reflectance profiles from liquid phantoms made up of absorbing dyes and fat emulsion scatterers over the spectral range of 450-1000 nm instantaneously. The hyperspectral reflectance data were analyzed by using a steady-state diffusion approximation model for semi-infinite homogeneous media. A calibration procedure was developed to compensate the nonuniform instrument response of the imaging system, and a curve-fitting algorithm was used to extract absorption and reduced scattering coefficients (mua and mus', respectively) for the phantoms in the wavelength range from 530 to 900 nm. The hyperspectral imaging system gave good measures of mua and mus' for the phantoms with average fitting errors of 12% and 7%, respectively. The hyperspectral imaging technique is fast, noncontact, and easy to use, which makes it especially suitable for measurement of the optical properties of turbid liquid and solid foods.

Time-gated optical system for depth-resolved functional brain imaging

We present a time-domain optical system for functional imaging of the adult head. We first describe the instrument, which is based on a Ti:Sapphire pulsed laser (wavelength 750-850 nm) and an intensified CCD camera enabling parallel detection of multiple fibers. We characterize the system in terms of sensitivity and signal-to-noise ratio, instrument response function, cross-talk, stability, and reproducibility. We then describe two applications of the instrument: the characterization of baseline optical properties of homogeneous scattering media, and functional brain imaging. For the second application, we developed a two-part probe consisting in two squares of 4 x 4 sources and 3 x 3 detectors. The laser source is time-multiplexed to define 4 states of 8 sources that can be turned on during the same camera frame while minimizing cross-talk. On the detection side, we use for each detector 7 fibers of different lengths creating an optical delay, and enabling simultaneous detection in 7 windows (by steps of 500 ps) for each detector. This multiple window detection allows depth sensitivity. The imaging probe was tested on dynamic phantoms and a preliminary result on an adult performing a motor task shows discrimination between superficial and cortical responses to the stimulus on both hemispheres.

Evaluation of anatomical structure and non-uniform distribution of imaging agent in near-infrared fluorescence-enhanced optical tomography

To date, the assessment of fluorescence-enhanced optical imaging has not been performed owing to (i) the lack of tools necessary for objective assessment of image quality (OAIQ), and (ii) the difficulty to test an untested diagnostic contrast agent in patient studies. Herein, we focus upon the development of a framework for OAIQ which includes a model to simulate both natural tissue heterogeneity as well as heterogeneous distribution of a molecularly targeted fluorophore. Specifically, we use a novel tomographic algorithm previously developed and validated in our laboratory (Roy and Sevick-Muraca, IEEE Trans. Med. Imaging, 2005). Our results show that image generation is (i) unaffected by normal anatomical heterogeneity manifested in endogenous tissue optical properties of absorption and scattering, and (ii) restricted by heterogeneous distribution of fluorophore in the background as the contrast is decreased.

Time-domain optical mammography SoftScan: initial results

RATIONALE AND OBJECTIVES

Near-infrared (NIR) technology appears promising as a noninvasive technique for breast cancer screening and diagnosis. The technology capitalizes on the relative transparency of human tissue in this spectral range and its sensitivity to the main components of the breast: water, lipid, and hemoglobin. In this study, the authors report quantitative measurements of these components and the functional contrast between healthy and diseased tissue.

MATERIALS AND METHODS

A four-wavelength time domain optical imaging system was used to perform noninvasive NIR measurements in the breast of 49 women both pre- and postmenopausal, ages 24-80. Algorithms based on a diffusive model of light transport provided absolute bulk and local values of breast constituent concentrations.

RESULTS

Important variations in the functional and structural NIR properties of the breast were observed. Demographics trend were noticed in accordance with breast physiology. In the 23 cases imaged with suspicious masses, the optical images were consistent with the mammographic findings. Substantial contrast between masses and adjacent tissue is observed. Moreover, consistent differences between malign and benign cases are found with optical imaging.

CONCLUSION

The results of this pilot study illustrate the sensitivity of optical techniques to the composition of the breast. In addition, preliminary data suggest that benign and malignant tumors can potentially be noninvasively differentiated with optical imaging. Moreover, statistically significant discrimination based on deoxy-hemoglobin content between malign and benign cases was found with optical imaging (P = .0184, one-tailed t test).

Simple peak shift analysis of time-of-flight data with a slow instrumental response function

Analysis of time-of-flight (TOF) data is sometimes limited by the instrumental response function, and optical parameters are extracted from the observed response curve by several mathematical methods, such as deconvolution. In contrast to this, we demonstrate that a method using shifts of the peak time of the response curve with different source-detector separations can yield the average path length of the light traveling in a tissue-like sample without deconvolution. In addition, combining the intensity information allows us to separate the scattering and absorption coefficients. This simple method is more robust in signal-to-noise ratio than the moment analysis, which also does not require the deconvolution procedure, because the peak position is not significantly dependent on the baseline fluctuation and the contamination of the scattering. The analysis is demonstrated by TOF measurements of an Intralipid solution at 800 nm, and is applied to the measurements at 1.29 microm, where the temporal response of photomultiplier tubes is not sufficiently good.

MR imaging of the her2/neu and 9.2.27 tumor antigens using immunospecific contrast agents

Molecular imaging of tumor antigens using immunospecific magnetic resonance (MR) contrast agents is a rapidly evolving field, which can potentially aid in early disease detection, monitoring of treatment efficacy, and drug development. In this study, we designed, synthetized, and tested in vitro two novel monocrystalline iron oxide nanoparticles (MION) conjugated to antibodies against the her2/neu tyrosine kinase receptor and the 9.2.27 proteoglycane sulfate. MION was synthetized by coprecipitation of iron II and iron III salts in 12-kD dextran solution; antibody coupling was performed by reductive amination. The relaxivity of the conjugates was 24.1-29.1 mM(-1) s(-1), with 1.8 to 2.1 antibody molecules per nanoparticle. A panel of cultured melanoma and mammary cell lines was used for testing. The cells were incubated with the particles at 16-32 microg Fe/ml in culture medium for 3 h at 37 degrees C, and investigated with immune fluorescence, transmission electron microscopy (TEM), MRI of cell suspensions in gelatine, and spectrophotometric iron determination. All receptor-positive cell lines, but not the controls, showed receptor-specific immune fluorescence, and strong changes in T(2) signal intensity at 1.5 T. The changes in 1/T(2) were between 1.5 and 4.6 s(-1) and correlated with the amount of cell-bound iron (R = 0.92). The relaxivity of cell-bound MION increased to 55.9 +/- 10.4 mM(-1) s(-1). TEM showed anti-9.2.27 conjugates binding to the plasma membrane, while the anti-her2/neu conjugates underwent receptor-mediated endocytosis. In conclusion, we obtained receptor-specific T(2) MR contrast with novel covalently bound, multivalent MION conjugates with anti-9.2.27 and anti-her2/neu to image tumor surface antigens. This concept can potentially be expanded to a large number of targets and to in vivo applications.

Catheter-based in vivo imaging of enzyme activity and gene expression: feasibility study in mice

PURPOSE

To construct and evaluate an interventional catheter-based imaging system for intravital monitoring of molecularly sensitive near-infrared fluorescent probes and optical marker genes.

MATERIALS AND METHODS

An imaging device that was based on a miniaturized fiberoptic sensor (MIFS) was built in which images created with a 2.7-F fiberoptic catheter were relayed through a dichroic mirror, through a bandpass filter, and on two independent cameras. This system permitted simultaneous recording of white-light and fluorescent images. Spatial resolution, spectral transmissions, and sensitivity were determined in vitro. In vivo testing was performed in nude mice bearing intraperitoneal tumors that express green fluorescent protein and in a mouse model of ovarian carcinoma with enzyme-activatable near-infrared probes sensitive to tumoral protease activity. Signal intensity on images of tumors and that on images of normal tissue were measured and compared with t test.

RESULTS

The catheter, which was advanced through an 18-gauge sheath, showed resolution of 7 line pairs per millimeter and detection limit for fluorochrome Cy5.5 of 1-10 pmol. Detection of endogenous green fluorescent protein gene expression was feasible in tumor nodules smaller than 1 mm in diameter (mean tumor signal intensity, 153.26 +/- 26.45 [SD], compared with that of adjacent nontumoral tissue of 36.73 +/- 11.69; P <.008). Similarly, activation of the near-infrared probe by tumoral proteases could be detected in peritoneal tumor seeds of ovarian cancer model with mean tumor signal intensity of 246.33 +/- 7.77 compared with that of adjacent nontumoral tissue of 41.56 +/- 18.64 (P <.001). Mean contrast-to-noise ratio in the near-infrared channel exceeded white-light contrast-to-noise ratio by a factor of 6.7 (P <.02).

CONCLUSION

With this system, in vivo MIFS imaging of gene expression, enzyme activity, and potentially other molecular events is feasible, through direct interventional access to several organs and body cavities and potentially through transvascular approaches.

An indeterministic Monte Carlo technique for fast time of flight photon transport through optically thick turbid media

A time-resolved indeterministic Monte Carlo (IMC) simulation technique is proposed for the efficient construction of the early part of the temporal point spread function (TPSF) of visible or near infrared photons transmitted through an optically thick scattering medium. By assuming a detected photon is a superposition of photon components, the photon is repropagated from a point in the original path where a significant delay in forward propagation occurred. A weight is then associated with each subsequently detected photon to compensate for shorter components. The technique is shown to reduce the computation time by a factor of at least 4 when simulating the sub-200 picosecond region of the TPSF and hence provides a useful tool for analysis of single photon detection in transillumination imaging.