Diffuse Optical Tomography Imaging Research Articles

Fast Hyperspectral Diffuse Optical Imaging Method with Joint Sparsity.

Diffuse optical tomography (DOT) is an important functional imaging modality in clinical diagnosis and treatment. As the number of wavelengths in the acquired DOT data grows, it becomes very challenging to reconstruct diffusion and absorption coefficients of tissue, i.e., a DOT image. In this paper, we consider the hyperspectral DOT (hyDOT) inverse problem as a multiple-measurement vector (MMV) problem by exploiting the joint sparsity of the images to be reconstructed. Then we propose a fast stochastic greedy algorithm based on the MMV stochastic gradient matching pursuit (MStoGradMP) and the mini-batching technique. Numerical results show that the proposed algorithm can achieve higher reconstruction accuracy with significantly reduced running time than the related gradient descent method with sparsity regularization.

Three-Dimensional Fluorescence Diffuse Optical Tomography Using the Adaptive Spatial Prior Approach

Traditionally, the reconstruction of fluorescence diffuse optical tomography (FDOT) images requires anatomical images as prior information. To integrate anatomical information about tumors or soft tissues into FDOT, additional equipment and procedures are required, involving co-registration, contrast-agent injection, and alignment. Herein, we propose the adaptive spatial prior approach for reconstructing FDOT images, implemented by using an initial guess of the tumor size and location for single and multi-tumor cases. This technique is most appropriately used when FDOT is combined with limited structural information or a different structural imaging system. This method iteratively modifies the location starting from the estimated location. We conduct simulations, phantom experiments and experiments on meat tissue using the adaptive spatial prior method. The results of these studies demonstrate the feasibility of the adaptive spatial prior approach for FDOT imaging.

Graph- and finite element-based total variation models for the inverse problem in diffuse optical tomography.

Total variation (TV) is a powerful regularization method that has been widely applied in different imaging applications, but is difficult to apply to diffuse optical tomography (DOT) image reconstruction (inverse problem) due to unstructured discretization of complex geometries, non-linearity of the data fitting and regularization terms, and non-differentiability of the regularization term. We develop several approaches to overcome these difficulties by: i) defining discrete differential operators for TV regularization using both finite element and graph representations; ii) developing an optimization algorithm based on the alternating direction method of multipliers (ADMM) for the non-differentiable and non-linear minimization problem; iii) investigating isotropic and anisotropic variants of TV regularization, and comparing their finite element (FEM)- and graph-based implementations. These approaches are evaluated on experiments on simulated data and real data acquired from a tissue phantom. Our results show that both FEM and graph-based TV regularization is able to accurately reconstruct both sparse and non-sparse distributions without the over-smoothing effect of Tikhonov regularization and the over-sparsifying effect of L1 regularization. The graph representation was found to out-perform the FEM method for low-resolution meshes, and the FEM method was found to be more accurate for high-resolution meshes.

Back-propagation neural network-based reconstruction algorithm for diffuse optical tomography.

.Diffuse optical tomography (DOT) is a promising noninvasive imaging modality and is capable of providing functional characteristics of biological tissue by quantifying optical parameters. The DOT image reconstruction is ill-posed and ill-conditioned, due to the highly diffusive nature of light propagation in biological tissues and limited boundary measurements. The widely used regularization technique for DOT image reconstruction is Tikhonov regularization, which tends to yield oversmoothed and low-quality images containing severe artifacts. It is necessary to accurately choose a regularization parameter for Tikhonov regularization. To overcome these limitations, we develop a noniterative reconstruction method, whereby optical properties are recovered based on a back-propagation neural network (BPNN). We train the parameters of BPNN before DOT image reconstruction based on a set of training data. DOT image reconstruction is achieved by implementing a single evaluation of the trained network. To demonstrate the performance of the proposed algorithm, we compare with the conventional Tikhonov regularization-based reconstruction method. The experimental results demonstrate that image quality and quantitative accuracy of reconstructed optical properties are significantly improved with the proposed algorithm.

Non-contact dynamic diffuse optical tomography imaging system for evaluating lower extremity vasculature.

A novel multi-view non-contact dynamic diffuse optical tomographic imaging system for the clinical evaluation of vasculature in the lower extremities is presented. The system design and implementation are described in detail, including methods for simultaneously obtaining and reconstructing diffusely reflected and transmitted light using a system of mirrors and a single CCD camera. The system and its performance using numeric simulations and optical phantoms. Measurements of a healthy foot in vivo demonstrates the potential of the system in assessing perfusion within the foot.

Fixed grid wavelet network segmentation on diffuse optical tomography image to detect sarcoma

Objective: To detect and explore the boundary of the sarcoma in Diffuse Optical Tomography (DOT) images, we need to extract the scattering and absorption property of the tissue at the cellular level. The DOT images suffer with lower optical resolution; therefore to improve the resolution in non-invasive imaging technique we apply Fixed Grid Wavelet Network (FGWN) image segmentation. Methods: We have subjected the reconstructed optical image to Vignette Correction to enhance the corners so that it traces the smooth boundary of tumor region. Fixed Grid Wavelet Network segmentation applied to reduce the training with the significant ortho-normal property. R, G and B values of optical image were considered as network inputs which lead to the formation of Wavelet network. Effective wavelet selection was based on Orthogonal Least Squares Algorithm and the network weights were calculated to optimize the network structure. The Mexican hat wavelet chosen facilitates the diffusion operator for image restoration, hence well-suited for Diffuse Optical Tomography (DOT) images.Results: Analysis made on data base of 30 DOT images and the 6 criteria results was evaluated. The boundary of the tumor region was traced on grayscale and the following Image Metrics were measured namely Mean Square Error, Root Mean Square Error, Peak Signal to Noise Ratio, Pearson Correlation Coefficient and Mean absolute error. The Receiver Operating Characteristics (ROC) was estimated at 99.527%, 88.73% and 93.8% with respect to sensitivity, specificity and overall accuracy. Conclusions: FGWN was compared with genetic algorithm and graph cut segmentation based on image metrics which exhibited 5.2% improvement and it was evaluated such that FGWN based image segmentation was superior to other methodologies.

Generalized curved beam back-projection method for near-infrared imaging using banana function.

Near-infrared imaging (NIRI) is a sub-surface imaging that makes a trade-off in recovery accuracy with depth of penetration. On the other hand, diffuse optical tomography (DOT) images tissue up to several centimeters. However, DOT reconstruction has a stability issue due to the inverse problem. This paper proposes a generalized continuous-wave technique to image objects of dimensions 4-6cm comparable to DOT. A nonlinear Rosenbrock's banana function is fitted to the approximate photon path, and the fit parameter thus obtained gives the penetration depth of each channel. The calculated values of absorption change are back-projected along these curved paths for reconstruction without solving the inverse problem. This function serves as an operator for image reconstruction. Here numerical simulations, experimental validation on wax phantom with inclusions, finger joint, and degraded apple have been performed to show potential of the proposed method in imaging. Thus this computationally efficient, reliable, and simple method is suitable for practical and real-time NIRI applications.

Modified CS-MUSIC for diffuse optical tomography using joint sparsity

In this paper, a modified compressive multiple signal classification (CS-MUSIC) algorithm has been proposed for diffuse optical tomography (DOT) reconstruction. The basic principle involved in DOT is that it illuminates the biological tissue using near infrared light and reconstructs the optical parameters of the tissue from the boundary measurements. The inverse problem of diffuse optical tomography is non-linear and severely ill-conditioned due to the zig-zag nature of light propagation by photons that diffuses through the tissue. Although nonlinear iterative methods are commonly used to solve this problem, they are computationally expensive since the forward problem has to be solved iteratively as well as they do not perform well for complex geometries. Recently, the DOT with compressive sensing (CS) has received a great attention due to its efficient possible reconstructions in DOT imaging. In this, the DOT inverse problem has been formulated as an multiple measurement vector (MMV) problem by using joint sparsity and CS frame work. The modified CS-MUSIC is a novel, non-iterative, and exact algorithm to reconstruct the absorption parameter change for Δα from the boundary data. In addition, this algorithm takes hybridization of sensor array signal processing and probabilistic compressive sensing. The experimental validation of the proposed algorithm has been done on a paraffin wax rectangular phantom through a DOT imaging setup. The performance metrics such as structural similarity index (SSIM), mean square error (MSE), normalized mean square error (NMSE) have been used to evaluate the performance of the reconstruction in this paper. Extensive numerical simulations show that the modified CS-MUSIC algorithm outperforms the current state-of-the-art algorithms and reliably reconstructs the absorption change in DOT.

Diffuse optical tomography using semiautomated coregistered ultrasound measurements.

Diffuse optical tomography (DOT) has demonstrated huge potential in breast cancer diagnosis and treatment monitoring. DOT image reconstruction guided by ultrasound (US) improves the diffused light localization and lesion reconstruction accuracy. However, DOT reconstruction depends on tumor geometry provided by coregistered US. Experienced operators can manually measure these lesion parameters; however, training and measurement time are needed. The wide clinical use of this technique depends on its robustness and faster imaging reconstruction capability. This article introduces a semiautomated procedure that automatically extracts lesion information from US images and incorporates it into the optical reconstruction. An adaptive threshold-based image segmentation is used to obtain tumor boundaries. For some US images, posterior shadow can extend to the chest wall and make the detection of deeper lesion boundary difficult. This problem can be solved using a Hough transform. The proposed procedure was validated from data of 20 patients. Optical reconstruction results using the proposed procedure were compared with those reconstructed using extracted tumor information from an experienced user. Mean optical absorption obtained from manual measurement was 0.21±0.06 cm-1 for malignant and 0.12±0.06 cm-1 for benign cases, whereas for the proposed method it was 0.24±0.08 cm-1 and 0.12±0.05 cm-1, respectively.

Effect of Shot Noise on Simultaneous Sensing in Frequency Division Multiplexed Diffuse Optical Tomographic Imaging Process.

Diffuse optical tomography (DOT) has been studied for use in the detection of breast cancer, cerebral oxygenation, and cognitive brain signals. As optical imaging studies have increased significantly, acquiring imaging data in real time has become increasingly important. We have developed frequency-division multiplexing (FDM) DOT systems to analyze their performance with respect to acquisition time and imaging quality, in comparison with the conventional time-division multiplexing (TDM) DOT. A large tomographic area of a cylindrical phantom 60 mm in diameter could be successfully reconstructed using both TDM DOT and FDM DOT systems. In our experiment with 6 source-detector (S-D) pairs, the TDM DOT and FDM DOT systems required 6.18 and 1 s, respectively, to obtain a single tomographic data set. While the absorption coefficient of the reconstruction image was underestimated in the case of the FDM DOT, we experimentally confirmed that the abnormal region can be clearly distinguished from the background phantom using both methods.

Greedy algorithms for diffuse optical tomography reconstruction

Diffuse optical tomography (DOT) is a noninvasive imaging modality that reconstructs the optical parameters of a highly scattering medium. However, the inverse problem of DOT is ill-posed and highly nonlinear due to the zig-zag propagation of photons that diffuses through the cross section of tissue. The conventional DOT imaging methods iteratively compute the solution of forward diffusion equation solver which makes the problem computationally expensive. Also, these methods fail when the geometry is complex. Recently, the theory of compressive sensing (CS) has received considerable attention because of its efficient use in biomedical imaging applications. The objective of this paper is to solve a given DOT inverse problem by using compressive sensing framework and various Greedy algorithms such as orthogonal matching pursuit (OMP), compressive sampling matching pursuit (CoSaMP), and stagewise orthogonal matching pursuit (StOMP), regularized orthogonal matching pursuit (ROMP) and simultaneous orthogonal matching pursuit (S-OMP) have been studied to reconstruct the change in the absorption parameter i.e, Δα from the boundary data. Also, the Greedy algorithms have been validated experimentally on a paraffin wax rectangular phantom through a well designed experimental set up. We also have studied the conventional DOT methods like least square method and truncated singular value decomposition (TSVD) for comparison. One of the main features of this work is the usage of less number of source–detector pairs, which can facilitate the use of DOT in routine applications of screening. The performance metrics such as mean square error (MSE), normalized mean square error (NMSE), structural similarity index (SSIM), and peak signal to noise ratio (PSNR) have been used to evaluate the performance of the algorithms mentioned in this paper. Extensive simulation results confirm that CS based DOT reconstruction outperforms the conventional DOT imaging methods in terms of computational efficiency. The main advantage of this study is that the forward diffusion equation solver need not be repeatedly solved.

Fast noninvasive functional diffuse optical tomography for brain imaging.

Advances in epilepsy studies have shown that specific changes in hemodynamics precede and accompany seizure onset and propagation. However, it has been challenging to noninvasively detect these changes in real time and in humans, due to the lack of fast functional neuroimaging tools. In this study, we present a functional diffuse optical tomography (DOT) method with the guidance of an anatomical human head atlas for 3-dimensionally mapping the brain in real time. Central to our DOT system is a human head interface coupled with a technique that can incorporate topological information of the brain surface into the DOT image reconstruction. The performance of the DOT system was tested by imaging motor tasks-involved brain activities on N = 6 subjects (3 epilepsy patients and 3 healthy controls). We observed diffuse areas of activations from the reconstructed [HbT] images of patients, relative to more focal activations for healthy subjects. Moreover, significant pretask hemodynamic activations were also seen in the motor cortex of patients, which indicated abnormal activities persistent in the brain of an epilepsy patient. This work demonstrates that fast functional DOT is a valuable tool for noninvasive 3-dimensional mapping of brain hemodynamics.

Kernel-based anatomically-aided diffuse optical tomography reconstruction

Image reconstruction in diffuse optical tomography (DOT) is a challenging task because its inverse problem is nonlinear, ill-posed, and ill-conditioned. Anatomical guidance from imaging modalities with high spatial resolution can substantially improve the quality of reconstructed DOT images. In this paper, inspired by the kernel methods in machine learning, we proposed a kernel method to introduce anatomical guidance into the DOT image reconstruction. In this kernel method, the optical absorption coefficient at each finite element node is represented as a function of a set of features obtained from anatomical images such as computed tomography (CT) images. Compared with Laplacian approaches that include structural priors, the proposed method does not require image segmentation. The proposed kernel method is validated with numerical simulations of 3D DOT reconstruction using synthetic CT data. 5% Gaussian noise was added to both the numerical DOT measurements and the simulated CT images. The proposed method was also validated by an agar phantom experiment with the anatomical guidance from a cone beam CT scan. The effects of voxel size and number of nearest neighbors in the kernel method on the reconstructed DOT images were studied. The results indicate that the spatial resolution and the accuracy of the reconstructed DOT images have been improved substantially after applying the anatomical guidance with the proposed kernel method comparing to the case without guidance. Furthermore, we demonstrated that the kernel method was able to utilize clinical breast CT images as anatomical guidance without segmentation. In addition, we found that the proposed kernel method was robust to the false positive guidance in the anatomical image.

Diffuse optical tomography for breast cancer imaging guided by computed tomography: A feasibility study.

Diffuse optical tomography (DOT) has attracted attentions in the last two decades due to its intrinsic sensitivity in imaging chromophores of tissues such as hemoglobin, water, and lipid. However, DOT has not been clinically accepted yet due to its low spatial resolution caused by strong optical scattering in tissues. Structural guidance provided by an anatomical imaging modality enhances the DOT imaging substantially. Here, we propose a computed tomography (CT) guided multispectral DOT imaging system for breast cancer imaging. To validate its feasibility, we have built a prototype DOT imaging system which consists of a laser at the wavelength of 650 nm and an electron multiplying charge coupled device (EMCCD) camera. We have validated the CT guided DOT reconstruction algorithms with numerical simulations and phantom experiments, in which different imaging setup parameters, such as projection number of measurements and width of measurement patch, have been investigated. Our results indicate that an air-cooling EMCCD camera is good enough for the transmission mode DOT imaging. We have also found that measurements at six angular projections are sufficient for DOT to reconstruct the optical targets with 2 and 4 times absorption contrast when the CT guidance is applied. Finally, we have described our future research plan on integration of a multispectral DOT imaging system into a breast CT scanner.

Photon fluencies for diffuse optical tomography systems

In this work, photon fluencies have been used to explore the depth of buried layers for diffuse optical tomography (DOT) systems. Back-reflected diffuse reflectance data were generated by Monte Carlo Modelling of Light Transport (MCMLT) simulations. Heterogeneous layers which have 1mm thickness were buried from 1mm to 8mm for 8 different simulations, consecutively. DOT systems are categorized based on the physical appearence as transmission through or back reflection geometry. Transmission through systems are more successful than back-reflected geometrical models from the view of the accuracy of the locations of the reconstructed images. On the other hand in the real world, back-reflection geometry is necessary for diagnostics and treatment purposes in the medical units and hospitals. DOT imaging systems basically use photon transport diffusion equation inside the turbid media. In literature, researchers and companies are using the linearized model of non-linear diffusion equation which is the first Taylor term of Rytov or Born approximation of radiative transport equation (RTE). The unknowns of the linearized equation system are differences of absorption coefficients over homogeneous background. After modeling of the linearized equation system, the next step is to solve the linearized equation system. Finally, researchers and company workers are using the forward model in different mathematical inverse problem solution algorithms which can be grouped as iterative, subspace, and regularization methods. Regularization methods are used very frequently hence the measurement system usually has noises and artifacts and the forward model transfer functions matrix is also not square. In this work a new method is presented for DOT imaging modalities without using any mathematical inverse problem solution algorithms. This new philosophy is using comparison of photon fluencies for different depth layers between homogeneous and heterogeneous tissue models. The new philosophical approach is being presented to explore the depth of the buried layers for DOT systems.

From brain to blood vessels and back: a noninvasive optical imaging approach.

The seminal work of Grinvald etal. has paved the way for the use of intrinsic optical signals measured with reflection methods for the analysis of brain function. Although this work has focused on the absorption signal associated with deoxygenation, due to its detailed mapping ability and good signal-to-noise ratio, Grinvald's group has also described other intrinsic signals related to increased blood flow, scattering effects directly related to neural activation, and pulsation effects related to arterial function. These intrinsic optical signals can also be measured using noninvasive diffuse optical topographic and tomographic imaging (DOT) methods that can be applied to humans. Here we compare the reflection and DOT methods and the evidence for each type of intrinsic signal in these two domains, with particular attention to work that has been conducted in our laboratory. This work reveals the refined two-way relationship that exists between vascular and neural phenomena in the brain: arterial health is related to normal brain structure and function, both across individuals and across brain regions within an individual, and neural function influences blood flow to specific cortical regions. DOT methods can provide quantitative tools for investigating these relationships in normal human subjects.

Capacitive charge approach to diffuse optical tomography systems

In this study the capacitive charge model approach has been explored to evaluate depth information of continuous wave (CW) diffuse optical tomography (DOT) systems. CW DOT systems have two geometric data acquisition structures depending on the laser source and photo-detector placements on the device surfaces. These geometries are transmission through and back reflection types. Most of the scanning geometries do not allow transmission through measurements. Hence, researchers had been trying to make more reliable back reflected CW DOT imaging systems for almost 30 years. Some researchers and companies had been using the mixture of transmission through and back reflection geometry models which is also equivalent to ring model. Ring model has been normally used for breast and newborn head imaging. The exponentially decaying nature of light gives less information about depth of scanning view for back reflected DOT devices. Diffusing light loses its energy along the depth profile. Back reflected escaping diffuse light involves less information about deeper voxels of interesting tissue or tissue like phantoms. On the other hand transmission through geometry uses only one depth surface thus giving more accurate information. In this work back reflected DOT geometry has been mimicked as a capacitor. Instead of electron accumulation versus time we used photon accumulation versus depth. The forward model transport matrix was built according to the time independent diffusion equation. One source and five different detector position’s experimental measurement data were acquired by multifiber spectrometers and simultenously acquired from homogenous and heterogenous multilayered, solid, sylindirical, silicon phantoms. Five different distance, heterogenous multilayer, detector's data have been divided over five homogenous detector's data to generate perturbation data. Perturbation data and the forward model weight matrix have been used to recover heterogenous, absorption differences from the homogenous medium and then two dimensional image has been constructed.

Comparing diffuse optical tomography and functional magnetic resonance imaging signals during a cognitive task: pilot study.

Diffuse optical tomography (DOT) measures concentration changes in both oxy- and deoxyhemoglobin providing three-dimensional images of local brain activations. A pilot study, which compares both DOT and functional magnetic resonance imaging (fMRI) volumes through t-maps given by canonical statistical parametric mapping (SPM) processing for both data modalities, is presented. The DOT series were processed using a method that is based on a Bayesian filter application on raw DOT data to remove physiological changes and minimum description length application index to select a number of singular values, which reduce the data dimensionality during image reconstruction and adaptation of DOT volume series to normalized standard space. Therefore, statistical analysis is performed with canonical SPM software in the same way as fMRI analysis is done, accepting DOT volumes as if they were fMRI volumes. The results show the reproducibility and ruggedness of the method to process DOT series on group analysis using cognitive paradigms on the prefrontal cortex. Difficulties such as the fact that scalp-brain distances vary between subjects or cerebral activations are difficult to reproduce due to strategies used by the subjects to solve arithmetic problems are considered. T-images given by fMRI and DOT volume series analyzed in SPM show that at the functional level, both DOT and fMRI measures detect the same areas, although DOT provides complementary information to fMRI signals about cerebral activity.

Diffuse optical tomography changes correlate with residual cancer burden after neoadjuvant chemotherapy in breast cancer patients.

Breast cancer (BC) patients who achieve a favorable residual cancer burden (RCB) after neoadjuvant chemotherapy (NACT) have an improved recurrence-free survival. Those who have an unfavorable RCB will have gone through months of ineffective chemotherapy. No ideal method exists to predict a favorable RCB early during NACT. Diffuse optical tomography (DOT) is a novel imaging modality that uses near-infrared light to assess hemoglobin concentrations within breast tumors. We hypothesized that the 2-week percent change in DOT-measured hemoglobin concentrations would associate with RCB. We conducted an observational study of 40 women with stage II-IIIC BC who received standard NACT. DOT imaging was performed at baseline and 2weeks after treatment initiation. We evaluated the associations between the RCB index (continuous measure), class (categorical 0, I, II, III), and response (RCB class 0/I=favorable, RCB class II/III=unfavorable) with changes in DOT-measured hemoglobin concentrations. The RCB index correlated significantly with the 2-week percent change in oxyhemoglobin [HbO2] (r=0.5, p=0.003), deoxyhemoglobin [Hb] (r=0.37, p=0.03), and total hemoglobin concentrations [HbT] (r=0.5, p=0.003). The RCB class and response significantly associated with the 2-week percent change in [HbO2] (p≤0.01) and [HbT] (p≤0.02). [HbT] 2-week percent change had sensitivity, specificity, positive, and negative predictive values for a favorable RCB response of 86.7, 68.4, 68.4, and 86.7%, respectively. The 2-week percent change in DOT-measured hemoglobin concentrations was associated with the RCB index, class, and response. DOT may help guide NACT for women with BC.

Computing Reduced Order Models via Inner-Outer Krylov Recycling in Diffuse Optical Tomography

In nonlinear imaging problems whose forward model is described by a partial differential equation (PDE), the main computational bottleneck in solving the inverse problem is the need to solve many large-scale discretized PDEs at each step of the optimization process. In the context of absorption imaging in diffuse optical tomography, de Sturler et al., [SIAM J. Sci. Compute., 37 (2015), pp. B495--B517] addres this bottleneck by using parametrized reduced order models, since the forward problem, typically posed in the frequency domain, is essentially the transfer function of a high-dimensional differential algebraic system. Although this approach drastically reduces the number of large linear systems to be solved, the construction of a candidate global basis for parametrized reduced models still requires the solution of many full order problems, the discretized PDE for multiple right-hand sides, and multiple parameter vectors. This step is followed by a rank-revealing factorization to compress the candidate...