Influence of fiber optic probe geometry on the applicability of inverse models of tissue reflectance spectroscopy: computational models and experimental measurements

Accurate recovery of tissue optical properties from in vivo spectral measurements is crucial for improving the clinical utility of optical spectroscopic techniques. The performance of inversion algorithms can be optimized for the specific fiber optic probe illumination-collection geometry. A diffusion-theory-based inversion method has been developed for the extraction of tissue optical properties from the shape of normalized tissue diffusion reflectance spectra, specifically tuned for a fiber probe that comprises seven hexagonally close-packed fibers. The central fiber of the probe goes to the spectrometer as the detecting fiber, and the surrounding six outer fibers are connected to the white-light source as illumination fibers. The accuracy of the diffusion-based inversion algorithm has been systematically assessed against Monte Carlo (MC) simulation as a function of probe geometry and tissue optical property combinations. By use of this algorithm, the spectral absorption and scattering coefficients of normal and cancerous tissue are efficiently retrieved. Although there are significant differences between the diffusion approximation and the MC simulation at short source-detector (SD) separations, we show that with our algorithm the tissue optical properties are well retrieved within the SD separation of 0.5-3 mm that is compatible with endoscopic specifications. The presented inversion method is computationally efficient for eventual real-time in vivo tissue diagnostics application.     

Optical tomographic image reconstruction from ultrafast time-sliced transmission measurements

Optical imaging and localization of objects inside a highly scattering medium, such as a tumor in the breast, is a challenging problem with many practical applications. Conventional imaging methods generally provide only two-dimensional (2-D) images of limited spatial resolution with little diagnostic ability. Here we present an inversion algorithm that uses time-resolved transillumination measurements in the form of a sequence of picosecond-duration intensity patterns of transmitted ultrashort light pulses to reconstruct three-dimensional (3-D) images of an absorbing object located inside a slab of a highly scattering medium. The experimental arrangement used a 3-mm-diameter collimated beam of 800-nm, 150-fs, 1-kHz repetition rate light pulses from a Ti:sapphire laser and amplifier system to illuminate one side of the slab sample. An ultrafast gated intensified camera system that provides a minimum FWHM gate width of 80 ps recorded the 2-D intensity patterns of the light transmitted through the opposite side of the slab. The gate position was varied in steps of 100 ps over a 5-ns range to obtain a sequence of 2-D transmitted light intensity patterns of both less-scattered and multiple-scattered light for image reconstruction. The inversion algorithm is based on the diffusion approximation of the radiative transfer theory for photon transport in a turbid medium. It uses a Green s function perturbative approach under the Rytov approximation and combines a 2-D matrix inversion with a one-dimensional Fourier-transform inversion to achieve speedy 3-D image reconstruction. In addition to the lateral position, the method provides information about the axial position of the object as well, whereas the 2-D reconstruction methods yield only lateral position.     

Green's function method in the radiative transfer problem. II. Spatially heterogeneous anisotropic surface

The most recent theoretical studies have shown that three-dimensional (3-D) radiation effects play an important role in the optical remote sensing of atmospheric aerosol and land surface reflectance. These effects may contribute notably to the error budget of retrievals in a broad range of sensor resolutions, introducing systematic biases in the land surface albedo data sets that emerge from the existing global observation systems. At the same time, 3-D effects are either inadequately addressed or completely ignored in data processing algorithms. Thus there is a need for further development of the radiative transfer theory that can rigorously treat both 3-D and surface anisotropy effects and yet be flexible enough to permit the development of fast forward and inversion algorithms. We describe a new theoretical solution to the 3-D radiative transfer problem with an arbitrary nonhomogeneous non-Lambertian surface. This solution is based on an exact semianalytical solution derived in operator form by the Green's function method. The numerical implementation is based on several parameterizations that accelerate the solution dramatically while keeping its accuracy within several percent under most general conditions.     

Semi-three-dimensional algorithm for time-resolved diffuse optical tomography by use of the generalized pulse spectrum technique

Although a foil three-dimensional (3-D) reconstruction with both 3-D forward and inverse models provide, the optimal solution for diffuse optical tomography (DOT), because of the 3-D nature of photon diffusion in tissue, it is computationally costly for both memory requirement and execution time in a conventional computing environment. Thus in practice there is motivation to develop an image reconstruction algorithm with dimensional reduction based on some modeling approximations. Here we have implemented a semi-3-D modified generalized pulse spectrum technique for time-resolved DOT, where a two-dimensional (2-D) distribution of optical properties is approximately assumed, while we retain 3-D distribution of photon migration in tissue. We have validated the proposed algorithm by reconstructing 3-D structural test objects from both numerically simulated and experimental date. We demonstrate our algorithm by comparing it with the calibrated 2-D reconstruction that is in widespread use as a shortcut to 3-D imaging and proving that the semi-3-D algorithm outperforms the calibrated 2-D algorithm.     

Multiwavelength three-dimensional near-infrared tomography of the breast: initial simulation,phantom, and clinical results

Three-dimensional (3D), multiwavelength near-infrared tomography has the potential to provide new physiological information about biological tissue function and pathological transformation. Fast and reliable measurements of multiwavelength data from multiple planes over a region of interest, together with adequate model-based nonlinear image reconstruction, form the major components of successful estimation of internal optical properties of the region. These images can then be used to examine the concentration of chromophores such as hemoglobin, deoxyhemoglobin, water, and lipids that in turn can serve to identify and characterize abnormalities located deep within the domain. We introduce and discuss a 3D modeling method and image reconstruction algorithm that is currently in place. Reconstructed images of optical properties are presented from simulated data, measured phantoms, and clinical data acquired from a breast cancer patient. It is shown that, with a relatively fast 3D inversion algorithm, useful images of optical absorption and scatter can be calculated with good separation and localization in all cases. It is also shown that, by use of the calculated optical absorption over a range of wavelengths, the oxygen saturation distribution of a tissue under investigation can be deduced from oxygenated and deoxygenated hemoglobin maps. With this method the reconstructed tumor from the breast cancer patient was found to have a higher oxy-deoxy hemoglobin concentration and also a higher oxygen saturation level than the background, indicating a ductal carcinoma that corresponds well to histology findings.     

Robust 3-D Motion Tracking From Stereo Images: A Model-Less Method

Traditional vision-based 3-D motion-estimation algorithms require given or calculated 3-D models while the motion is being tracked. We propose a high-speed extended-Kalman-filter-based approach that recovers camera position and orientation from stereo image sequences without prior knowledge, as well as the procedure for the reconstruction of 3-D structures. Empowered by the use of a trifocal tensor, the computation step of 3-D models can be eliminated. The algorithm is thus flexible and can be applied to a wide range of domains. The twist motion model is also adopted to parameterize the 3-D motion. It is minimal since it only has six parameters as opposed to seven parameters in quaternion and 12 parameters in matrix representation. The motion representation is robust because it does not suffer from singularities as Euler angles. Due to the fact that the number of parameters to be estimated is reduced, our algorithm is more efficient, stable, and accurate than traditional approaches. The proposed method has been applied to recover the motion from stereo image sequences taken by a robot and a handheld stereo rig. The results are accurate compared to the ground truths. It is shown in the experiment that our algorithm is not susceptible to outlying point features with the application of a validation gate.     

Practical fully three-dimensional reconstruction algorithms for diffuse optical tomography

We have developed an efficient fully three-dimensional (3D) reconstruction algorithm for diffuse optical tomography (DOT). The 3D DOT, a severely ill-posed problem, is tackled through a pseudodynamic (PD) approach wherein an ordinary differential equation representing the evolution of the solution on pseudotime is integrated that bypasses an explicit inversion of the associated, ill-conditioned system matrix. One of the most computationally expensive parts of the iterative DOT algorithm, the reevaluation of the Jacobian in each of the iterations, is avoided by using the adjoint-Broyden update formula to provide low rank updates to the Jacobian. In addition, wherever feasible, we have also made the algorithm efficient by integrating along the quadratic path provided by the perturbation equation containing the Hessian. These algorithms are then proven by reconstruction, using simulated and experimental data and verifying the PD results with those from the popular Gauss-Newton scheme. The major findings of this work are as follows: (i) the PD reconstructions are comparatively artifact free, providing superior absorption coefficient maps in terms of quantitative accuracy and contrast recovery; (ii) the scaling of computation time with the dimension of the measurement set is much less steep with the Jacobian update formula in place than without it; and (iii) an increase in the data dimension, even though it renders the reconstruction problem less ill conditioned and thus provides relatively artifact-free reconstructions, does not necessarily provide better contrast property recovery. For the latter, one should also take care to uniformly distribute the measurement points, avoiding regions close to the source so that the relative strength of the derivatives for measurements away from the source does not become insignificant.     

Performance evaluation of iterative tomography algorithms for incomplete projection data

Projection data obtained through optical techniques for tomographic measurements, such as interferometry for refractive-index-based measurements, are often incomplete. This is due to limitations in the optical system, data storage, and alignment and vignette issues. Algebraic iterative reconstruction techniques are usually favored for such incomplete projections. A number of iterative algorithms, based on additive and multiplicative corrections, are used with a known simulated phantom and noise source to assess the reconstruction performance of incomplete data sets. In addition, we present reconstructions using experimental data obtained from a coherent gradient sensing interferometer for a steady temperature field in a fluid medium. We tested the algorithms using the simulated data set for incompleteness conditions similar to those found in the experimental data, and the best-performing algorithm is identified.     

Three-dimensional simulation of near-infrared diffusion in tissue: boundary condition and geometry analysis for finite-element image reconstruction

Imaging of tissue with near-infrared spectral tomography is emerging as a practicable method to map hemoglobin concentrations within tissue. However, the accurate recovery of images by using modeling methods requires a good match between experiments and the model prediction of light transport in tissue. We illustrate the potential for a match between (i) three-dimensional (3-D) frequency-domain diffusion theory, (ii) two-dimensional diffusion theory, (iii) Monte Carlo simulations, and (iv) experimental data from tissue-simulating phantoms. Robin-type boundary conditions are imposed in the 3-D model, which can be implemented with a scalar coupling coefficient relating the flux through the surface to the diffuse fluence rate at the same location. A comparison of 3-D mesh geometries for breast imaging indicates that relative measurements are sufficiently similar when calculated on either cylindrical or female breast shapes, suggesting that accurate reconstruction may be achieved with the simpler cylindrical mesh. Simulation studies directly assess the effects from objects extending out of the image plane, with results suggesting that spherically shaped objects reconstruct at lower contrast when their diameters are less than 15-20 mm. The algorithm presented here illustrates that a 3-D forward diffusion model can be used with circular tomographic measurements to reconstruct two-dimensional images of the interior absorption coefficient.     

Full- and minimal-scan reconstruction algorithms for fan-beam diffraction tomography

Diffraction tomography (DT) is a tomographic inversion technique that reconstructs the spatially variant refractive-index distribution of a scattering object. In fan-beam DT, the interrogating radiation is not a plane wave but rather a cylindrical wave front emanating from a line source located a finite distance from the scattering object. We reveal and examine the redundant information that is inherent in the fan-beam DT data function. Such redundant information can be exploited to reduce the reconstructed image variance or, alternatively, to reduce the angular scanning requirements of the fan-beam DT experiment. We develop novel filtered backpropagation and estimate-combination reconstruction algorithms for full-scan and minimal-scan fan-beam DT. The full-scan algorithms utilize measurements taken over the angular range 0 相似文献   

Real-time monitoring and growth control of Si-gradient-index structures by multiwavelength ellipsometry

Real-time monitoring and control of the growth of plasma-deposited gradient-index structures by multiwavelength phase-modulated ellipsometry are presented. An efficient method for estimating the optical parameters based on a least-squares fitting on the most recent recorded measurements is described. This method is used for real-time monitoring of the outerlayer refractive index and rate of deposition during deposition of inhomogeneous transparent films. An accurate integral expansion of the reflection coefficient, giving a continuous optical model describing inhomogeneous films, is used in the real-time modeling of the deposited structure. A variety of increasing complexities of the optical model is studied within inversion algorithms. Furthermore several ways of defining the optical parameters to be estimated are discussed. Inversion of simulated growth trajectories by using the described algorithms show what kind of information is available from the various approximations and in what conditions they are useful. Since real-time measurements during growth involves statistical noise and systematic errors, it becomes necessary to stabilize the fit. Various stabilizing functionals are discussed and implemented to solve this problem. Several plasma-deposited silicon oxynitride gradient-index structures where both rate of deposition and the refractive index are varied continuously are inverted in real time. As an example of application, a successful real-time control of the growth of a linear gradient index is demonstrated by using inversion algorithms. Inversion algorithms are extremely fast, with calculation times from less than a second (for the lowest-order approximation) to ~3 s.     

一种新的频谱分析图像重建算法及模拟

这种频谱分析图像重建算法利用二维傅里叶级数的各正交基函数的投影向量,去逼近图像 函数的投影向量,利用优化准则,得到各正交基函数的权系数,再经过傅里叶反变换,即可重建出原始的图像函数。计算机模拟表明,该算法重建时间不足1秒,选择合适的近似级数,对复杂场的图像重建,也能使最大误差不超过5%,平均误差不超过1%。     

Spatially variant regularization improves diffuse optical tomography

Diffuse tomography with near-infrared light has biomedical application for imaging hemoglobin, water, lipids, cytochromes, or exogenous contrast agents and is being investigated for breast cancer diagnosis. A Newton-Raphson inversion algorithm is used for image reconstruction of tissue optical absorption and transport scattering coefficients from frequency-domain measurements of modulated phase shift and light intensity. A variant of Tikhonov regularization is examined in which radial variation is allowed in the value of the regularization parameter. This method minimizes high-frequency noise in the reconstructed image near the source-detector locations and can produce constant image resolution and contrast across the image field.     

Vertical profiles of microphysical particle properties derived from inversion with two-dimensional regularization of multiwavelength Raman lidar data: experiment

Inversion with two-dimensional (2-D) regularization is a new methodology that can be used for the retrieval of profiles of microphysical properties, e.g., effective radius and complex refractive index of atmospheric particles from complete (or sections) of profiles of optical particle properties. The optical profiles are acquired with multiwavelength Raman lidar. Previous simulations with synthetic data have shown advantages in terms of retrieval accuracy compared to our so-called classical one-dimensional (1-D) regularization, which is a method mostly used in the European Aerosol Research Lidar Network (EARLINET). The 1-D regularization suffers from flaws such as retrieval accuracy, speed, and ability for error analysis. In this contribution, we test for the first time the performance of the new 2-D regularization algorithm on the basis of experimental data. We measured with lidar an aged biomass-burning plume over West/Central Europe. For comparison, we use particle in situ data taken in the smoke plume during research aircraft flights upwind of the lidar. We find good agreement for effective radius and volume, surface-area, and number concentrations. The retrieved complex refractive index on average is lower than what we find from the in situ observations. Accordingly, the single-scattering albedo that we obtain from the inversion is higher than what we obtain from the aircraft data. In view of the difficult measurement situation, i.e., the large spatial and temporal distances between aircraft and lidar measurements, this test of our new inversion methodology is satisfactory.     

Imaging of fluorescent yield and lifetime from multiply scattered light reemitted from random media

The feasibility of employing fluorescent contrast agents to perform optical imaging in tissues and other scattering media has been examined through computational studies. Fluorescence lifetime and yield can give crucial information about local metabolite concentrations or environmental conditions within tissues. This information can be employed toward disease detection, diagnosis, and treatment if noninvasively quantitated from reemitted optical signals. However, the problem of inverse image reconstruction of fluorescence yield and lifetime is complicated because of the highly scattering nature of the tissue. Here a light propagation model employing the diffusion equation is used to account for the scattering of both the excitation and fluorescent light. Simulated measurements of frequency-domain parameters of fluorescent modulated ac amplitude and phase lag are used as inputs to an inverse image-reconstruction algorithm, which employs the diffusion model to predict frequency-domain measurements resulting from a modulated input at the phantom periphery. In the inverse image-reconstruction algorithm, a Newton-Raphson technique combined with a Marquardt algorithm is employed to converge on the fluorescent properties within the medium. The successful reconstruction of both the fluorescence yield and lifetime in the case of a heterogeneous fluorophore distribution within a scattering medium has been demonstrated without a priori information or without the necessity of obtaining absence images.     

Numerically robust minimal‐scan reconstruction algorithms for diffraction tomography via radon transform inversion

It is widely believed that measurements from a full angular range of 2π are generally required to exactly reconstruct a complex‐valued refractive index distribution in diffraction tomography (DT). In this work, we developed a new class of minimal‐scan reconstruction algorithms for DT that utilizes measurements only over the angular range 0 ≤ ? ≤ 3π/2 to perform an exact reconstruction. These algorithms, referred to as minimal‐scan estimate‐combination (MS‐E‐C) reconstruction algorithms, effectively operate by transforming the DT reconstruction problem into a conventional x‐ray CT reconstruction problem that requires inversion of the Radon transform. We performed computer simulations to compare the noise and numerical properties of the MS‐E‐C algorithms against existing filtered backpropagation‐based algorithms. © 2002 Wiley Periodicals, Inc. Int J Imaging Syst Technol 12, 84–91, 2002; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ima.10014     

3-D ultrasound volume reconstruction using the direct frame interpolation method

A new method for 3-D ultrasound volume reconstruction using tracked freehand 3-D ultrasound is proposed. The method is based on solving the forward volume reconstruction problem using direct interpolation of high-resolution ultrasound B-mode image frames. A series of ultrasound B-mode image frames (an image series) is acquired using the freehand scanning technique and position sensing via optical tracking equipment. The proposed algorithm creates additional intermediate image frames by directly interpolating between two or more adjacent image frames of the original image series. The target volume is filled using the original frames in combination with the additionally constructed frames. Compared with conventional volume reconstruction methods, no additional filling of empty voxels or holes within the volume is required, because the whole extent of the volume is defined by the arrangement of the original and the additionally constructed B-mode image frames. The proposed direct frame interpolation (DFI) method was tested on two different data sets acquired while scanning the head and neck region of different patients. The first data set consisted of eight B-mode 2-D frame sets acquired under optimal laboratory conditions. The second data set consisted of 73 image series acquired during a clinical study. Sample volumes were reconstructed for all 81 image series using the proposed DFI method with four different interpolation orders, as well as with the pixel nearest-neighbor method using three different interpolation neighborhoods. In addition, volumes based on a reduced number of image frames were reconstructed for comparison of the different methods' accuracy and robustness in reconstructing image data that lies between the original image frames. The DFI method is based on a forward approach making use of a priori information about the position and shape of the B-mode image frames (e.g., masking information) to optimize the reconstruction procedure and to reduce computation times and memory requirements. The method is straightforward, independent of additional input or parameters, and uses the high-resolution B-mode image frames instead of usually lower-resolution voxel information for interpolation. The DFI method can be considered as a valuable alternative to conventional 3-D ultrasound reconstruction methods based on pixel or voxel nearest-neighbor approaches, offering better quality and competitive reconstruction time.     

Investigating the enhancement of three-dimensional diffraction tomography by using multiple illumination planes

The three dimensional (3-D) extension of the two well-known diffraction tomography algorithms, namely, direct Fourier interpolation (DFI) and filtered backpropagation (FBP), are presented and the problem of the data needed for a full 3-D reconstruction is investigated. These algorithms can be used efficiently to solve the inverse scattering problem for weak scatterers in the frequency domain under the first-order Born and Rytov approximations. Previous attempts of 3-D reconstruction with plane-wave illumination have used data obtained with the incident direction restricted at the xy plane. However, we show that this restriction results in the omission of the contribution of certain spatial frequencies near the omegaz axis for the final reconstruction. The effect of this omission is studied by comparing the results of reconstruction with and without data obtained from other incident directions that fill the spatial frequency domain. We conclude that the use of data obtained for incident direction in only the xy plane is sufficient to achieve a satisfactory quality of reconstruction for a class of objects presenting smooth variation along the z axis, while abrupt variations along the z axis cannot be imaged. This result should be taken into account in the process of designing the acquisition geometry of a tomography scanner.     

Enumeration algorithm for determination of binding constants in capillary electrophoresis

With more accurate simulation models and more efficient algorithms becoming available, the binding constants of an affinity interaction can be obtained from much simpler experiments using capillary electrophoresis. With the enumeration algorithm, all possible combinations of the binding constant and the complex mobility in certain ranges that could result in the experimental migration time of an injected analyte are extracted from a 3-D surface, which depicts the migration times resulting from different values of the binding constant and the mobility of the complex formed between the interacting pair, to form a 2-D curve. When the experimental conditions are changed, the analyte migration time will also change. A new 2-D curve can be constructed from another 3-D surface on the basis of the pairs of binding constants and complex mobility values that could result in the new migration time. Because the true binding constant and complex mobility values have to be the same for both experimental conditions under the same temperature, there has to be a point where both 2-D curves will converge. The coordinates of the converging point give the values for a binding constant and a complex mobility that will fit all 2-D curves generated under certain experimental conditions. p-Nitrophenol is used as the analyte, beta-cyclodextrin is used as the additive, and a one-cell model is used to simulate affinity CE. The experimental conditions that can improve the accuracy of the binding constants are discussed.     

Comparison of the filtered backpropagation and the filtered backprojection algorithms for quantitative tomography

We compare the filtered backpropagation algorithm with the filtered backprojection algorithm for reconstructing the complex refractive-index distribution of semitransparent, cylindrical objects. Before reconstruction, the recorded scattered light is propagated back to the reconstruction area by inverse diffraction. Our comparison is based on computer-simulated data, and experimental optical data obtained from fibers with step-index, graded-index, and uniform-index distributions. The results show that both the filtered backpropagation algorithm and the filtered backprojection algorithm can produce accurate reconstructions of the complex refractive-index distribution as long as the weak-scattering approximation is valid. The good agreement between the results obtained from these two reconstruction algorithms indicates that the errors introduced by the assumption of straight-line propagation inside the object are negligible compared with those introduced by the weak-scattering approximation.