Multi-energy Images Research Articles

Spectral CT Reconstruction—ASSIST: Aided by Self-Similarity in Image-Spectral Tensors

Spectral computed tomography (CT) reconstructs multienergy images from data in different energy bins. However, these reconstructed images can be contaminated by noise due to the limited numbers of photons in the corresponding energy bins. In this paper, we propose a spectral CT reconstruction method aided by self-similarity in image-spectral tensors, which utilizes the self-similarity of patches in both spatial and spectral domains. Patches with similar structures identified by a joint spatial and spectral searching strategy form a basic tensor unit, and can be utilized to improve image quality. Specifically, each tensor is decomposed into a low-rank component and a sparse component, which respectively represent the stable structures and feature differences across different energy bins. The augmented Lagrange method is applied to optimize the proposed objective function. To validate the performance of the proposed method, several simulated clinical and real data experiments are performed. The qualitative and quantitative results demonstrate that the proposed method outperforms several representative state-of-the-art algorithms in terms of preserving image details and reducing artifacts.

Dynamic-dual-energy spectral CT for improving multi-material decomposition in image-domain

Dual-energy CT, as well as spectral CT, has a great potential in material decomposition. However, dual-energy CT is difficult to apply to multi-material decomposition because the number of energy bins is limited to two. Current spectral CT systems have more energy bins, but the statistical noise in each energy bin is high because of the decreased photon number, which causes errors in the material decomposition results. In this paper, we propose a dynamic-dual-energy spectral CT for accurate multi-material decomposition. In the course of scanning, the energy threshold of the dynamic-dual-energy detector randomly changes to obtain the spectral information of photons. With the proposed statistical noise-weighted tPRISM algorithm, the multi-energy image reconstruction using dynamic-dual-energy CT data was implemented, followed by multi-material decomposition. Both simulation and experiment results show that the multi-energy reconstruction and multi-material decomposition using the dynamic-dual-energy method are more accurate and have less noise compared with that of the conventional static-multi-energy method with the same number of energy bins. The ring artifacts which are severe in the experimental data simulation and experiment results using the conventional spectral CT method are reduced in great extent when using our proposed method. In conclusion, our proposed dynamic-dual-energy spectral CT method is highly feasible and has a great potential in high-quality multi-material decomposition.

Unsupervised learning methods in X-ray spectral imaging material segmentation

In this work, we have investigated a number of unsupervised learning methods for material segmentation in projection x-ray imaging with a spectral detector. A phantom containing two hard materials (glass, steel) and three soft materials (PVC, polypropylene, and PFTE) all embedded in PMMA was imaged with a 5 energy bin spectal detector. The projection images were utilized to test nine unsupervised learning algorithms for automated material segmentation. Each algorithm was investigated using single energy (SE), dual energy (DE) and multi energy (ME) images. Clustering results were scored based on homogeneity and completeness of the clusters, which were combined into the Rosenberg and Hirshberg's V-measure. Principle component analysis (PCA), independent component analysis (ICA), and non-negative matrix factorization (NMF) were tested as dimensional reduction methods. ME, DE and SE material segmentation was performed using five, two, and single energy images, respectively. ME had the highest V-measure on the soft materials using PCA and a novel interpolating bayesian gaussian mixture model (BGMM) clustering with a V-measure of 0.71. This was by 3.5% better than DE and 20.3% better than SE. Conversely, SE imaging was most capable of hard tissue segmentation using the standard BGMM, with a V-measures of 0.84. This was 6.3% better than DE and 5.0% better than ME. This work demonstrated that ME x-ray imaging might be superior in segmenting soft tissues compared to conventional SE x-ray imaging.

Clinical Utility of Multi-Energy Spectral Photon-Counting Computed Tomography in Crystal Arthritis.

To determine whether novel multi-energy spectral photon-counting computed tomography (SPCCT) imaging can detect and differentiate between monosodium urate (MSU), calcium pyrophosphate (CPP), and hydroxyapatite (HA) crystal deposits ex vivo. A finger with a subcutaneous gouty tophus and a calcified knee meniscus excised at the time of surgery were obtained. The finger was imaged using plain x-ray, dual-energy CT (DECT), and multi-energy SPCCT. Plain x-ray and multi-energy SPCCT images of the meniscus were acquired. For validation purposes, samples of the crystals were obtained from the tophus and meniscus, and examined by polarized light microscopy and/or x-ray diffraction. As further validation, synthetic crystal suspensions of MSU, CPP, and HA were scanned using multi-energy SPCCT. Plain x-ray of the gouty finger revealed bone erosions with overhanging edges. DECT and multi-energy SPCCT both showed MSU crystal deposits; SPCCT was able to show finer detail. Plain x-ray of the calcified meniscus showed chondrocalcinosis consistent with CPP, while SPCCT showed and differentiated CPP and HA. Multi-energy SPCCT can not only detect, differentiate, and quantify MSU crystal deposits in a gouty finger ex vivo, but also specifically detect, identify, and quantify CPP within an osteoarthritic meniscus, and distinguish them from HA crystal deposits. There is potential for multi-energy SPCCT to become useful in the diagnosis of crystal arthropathies.

Improving iodine contrast to noise ratio using virtual monoenergetic imaging and prior-knowledge-aware iterative denoising (mono-PKAID)

Multi-energy CT acquires simultaneous multiple x-ray attenuation measurements from different energy spectra which facilitates the computation of virtual monoenergetic images (VMI) at a specific photon energy (keV). Since the contrast between iodine attenuation and the attenuation of surrounding soft tissues increases at lower x-ray energies, VMIs in the range of 40–70 keV can be used to improve iodine visualization. However, at lower energy levels, image noise in VMIs is substantially increased, which counteracts the benefits from the increased iodine contrast, resulting in a decreased iodine contrast-to-noise ratio (CNR). There exists considerable data redundancy between multi-energy CT images created from the same acquisition. Similarly, a substantial spatio-spectral data redundancy exists between multi-energy CT images and the corresponding VMIs. In this work, we develop a denoising framework that exploits this data redundancy to improve iodine CNR in the VMIs. We accomplish this by applying prior-knowledge-aware iterative denoising to low-energy VMIs; we refer to the denoised images as mono-PKAID images. The proposed framework was evaluated using phantom and in vivo data acquired on a research whole-body photon-counting-detector CT, as well as using data from a commercial dual-source dual-energy CT system. The results of phantom experiments show that the proposed framework can preserve image resolution and noise texture compared to the original VMIs, while reducing noise to improve iodine CNR. Quantitative measurements show that the iodine CNR of 50 keV VMI is improved by 1.8-fold using the proposed method, relative to the VMI produced using commercial software (Mono+). With mono-PKAID, VMIs at lower keV take full advantage of higher iodine contrast without substantially increasing image noise. These observations were confirmed using patient data sets, which demonstrated that mono-PKAID reduced image noise, improved CNR in anatomical regions with iodine perfusion by 1.8-fold, and potentially enhanced the visibility of focal liver lesions.

Total image constrained diffusion tensor for spectral computed tomography reconstruction

• The TICDT is derived with the supplementary information from a high-quality total image. • TICDT can maintain structural features of total image even when target image is highly corrupted. • An alternating minimization algorithm is adapted to solve the PWLS-TICDT reconstruction model. • This is the first time that diffusion tensor is applied to spectral CT reconstruction. Photon counting detector (PCD)-based spectral computed tomography (CT) is a promising imaging technique that enables high energy resolution imaging with narrow energy bins. However, the image quality is degraded because the number of photons in each energy bin is less than the number of photons in the full spectrum. To reconstruct high quality spectral CT images with narrow energy bins, we developed a total image constrained diffusion tensor (TICDT) for statistical iterative reconstruction (SIR) based on a penalized weighted least-squares (PWLS) principle, which is called “PWLS-TICDT.” Specifically, TICDT uses supplementary information from a high-quality total image as a structural prior for SIR, so that the narrow energy bin image can be enhanced, while some primary features are preserved. We also developed an alternating minimization algorithm to solve the associated objective function. We conducted qualitative and quantitative studies to validate and evaluate the PWLS-TICDT method using digital phantoms and preclinical data. Results from both numerical simulation and real PCD data studies show that the proposed PWLS-TICDT method achieves noticeable gains over competing methods in terms of suppressing noise, detecting low contrast objects, and preserving resolution. More importantly, the multi-energy images reconstructed by PWLS-TICDT method can generate more accurate basis material decomposition results than the other methods.

Nonlocal low-rank and sparse matrix decomposition for spectral CT reconstruction

Spectral computed tomography (CT) has been a promising technique in research and clinics because of its ability to produce improved energy resolution images with narrow energy bins. However, the narrow energy bin image is often affected by serious quantum noise because of the limited number of photons used in the corresponding energy bin. To address this problem, we present an iterative reconstruction method for spectral CT using nonlocal low-rank and sparse matrix decomposition (NLSMD), which exploits the self-similarity of patches that are collected in multi-energy images. Specifically, each set of patches can be decomposed into a low-rank component and a sparse component, and the low-rank component represents the stationary background over different energy bins, while the sparse component represents the rest of the different spectral features in individual energy bins. Subsequently, an effective alternating optimization algorithm was developed to minimize the associated objective function. To validate and evaluate the NLSMD method, qualitative and quantitative studies were conducted by using simulated and real spectral CT data. Experimental results show that the NLSMD method improves spectral CT images in terms of noise reduction, artifact suppression and resolution preservation.

Superiorization-based multi-energy CT image reconstruction

The recently-developed superiorization approach is efficient and robust for solving various constrained optimization problems. This methodology can be applied to multi-energy CT image reconstruction with the regularization in terms of the prior rank, intensity and sparsity model (PRISM). In this paper, we propose a superiorized version of the simultaneous algebraic reconstruction technique (SART) based on the PRISM model. Then, we compare the proposed superiorized algorithm with the Split-Bregman algorithm in numerical experiments. The results show that both the Superiorized-SART and the Split-Bregman algorithms generate good results with weak noise and reduced artefacts.

Material decomposition with the multi-energy attenuation coefficient ratio by using a multiple discriminant analysis

The objective of this study was to develop a spectral CT system using a photon counting detector and to decompose materials by applying a multiple discriminant analysis (MDA) to the energy-dependent attenuation coefficient ratios. We imaged cylindrical phantoms of Polymethyl methacrylate (PMMA) with four holes filled with calcium chloride, iodine, and gold nanoparticle contrast agents. The attenuation coefficients were measured via reconstructed multi-energy images, and the linear attenuation ratio was used for material identification. The MDA projection matrix, determined from training phantoms, was used to identify the four materials in the testing phantoms. For quantification purposes, the relationships between the attenuation coefficients at multiple energy bins and the concentrations were characterized by using the least-squares method for each material. The mean identification accuracy for each of the three materials were 0.94 ± 0.09 for iodine, 0.96 ± 0.07 for gold nanoparticles, and 0.92 ± 0.05 for calcium chloride. The mean quantification errors were 1.90 ± 1.58% for iodine, 3.85 ± 3.13% for gold nanoparticle, and 3.40 ± 2.62% for calcium chloride. The developed multi-energy CT system based on the photon-counting detector with MDA can precisely decompose the four materials.

Soft tissue imaging with photon counting spectroscopic CT

The purpose of this work was experimental investigation of photon counting spectroscopic CT (PCS-CT) imaging of anatomical soft tissue with clinically relevant size. The imaging experiments were performed using a spectroscopic CT system based on CdZnTe photon counting detector with two rows of pixels, 256 pixels in each row, 1 × 1 mm2 pixel size, and 25.6 cm detector length. The detector could split the x-ray energy spectrum to 5 regions (energy bins), and acquire 5 multi-energy (spectroscopic) CT images in a single CT scan. A sample of round shaped anatomical soft tissue of 14 cm diameter including lean and fat was used for imaging. To avoid the negative effect of anatomical noise on quantitative analysis, a spectroscopic CT phantom with tissue equivalent solid materials was used. The images were acquired at 60, 90, and 120 kVp tube voltages, and spectroscopic image series were acquired with 3 and 5 energy bins. Spectroscopic CT numbers were introduced and used to evaluate an energy selective image series. The anatomical soft tissue with 14 cm diameter was visualized with good quality and without substantial artifacts by the photon counting spectroscopic CT system. The effects of the energy bin crosstalk on spectroscopic CT numbers were quantified and analyzed. The single and double slice PCS-CT images were acquired and compared. Several new findings were observed, including the effect of soft tissue non-uniformity on image artifacts, unique status of highest energy bin, and material dependent visualization in spectroscopic image series. Fat–lean decomposition was performed using dual energy subtraction and threshold segmentation methods, and compared. Using K-edge filtered x-rays improved fat–lean decomposition as compared to conventional x-rays. Several new and important aspects of the PCS-CT were investigated. These include imaging soft tissue with clinically relevant size, single- and double-slice PCS-CT imaging, using spectroscopic CT numbers to quantify multi-energy PCS-CT images, application of K-edge filtered x-rays for improved soft tissue decomposition, and several others. The study suggests that the presented PCS-CT technology meets the requirements of a particular clinical application, i.e. dedicated breast CT.

TU-F-18A-04: Use of An Image-Based Material-Decomposition Algorithm for Multi-Energy CT to Determine Basis Material Densities

Purpose: Published methods for image-based material decomposition with multi-energy CT images have required the assumption of volume conservation or accurate knowledge of the x-ray spectra and detector response. The purpose of this work was to develop an image-based material-decomposition algorithm that can overcome these limitations. Methods: An image-based material decomposition algorithm was developed that requires only mass conservation (rather than volume conservation). With this method, using multi-energy CT measurements made with n=4 energy bins, the mass density of each basis material and of the mixture can be determined without knowledge of the tube spectra and detector response. A digital phantom containing 12 samples of mixtures from water, calcium, iron, and iodine was used in the simulation (Siemens DRASIM). The calibration was performed by using pure materials at each energy bin. The accuracy of the technique was evaluated in noise-free and noisy data under the assumption of an ideal photon-counting detector. Results: Basis material densities can be estimated accurately by either theoretic calculation or calibration with known pure materials. The calibration approach requires no prior information about the spectra and detector response. Regression analysis of theoretical values versus estimated values results in excellent agreement for both noise-free and noisy data. For the calibration approach, the R-square values are 0.9960+/−0.0025 and 0.9476+/−0.0363 for noise-free and noisy data, respectively. Conclusion: From multi-energy CT images with n=4 energy bins, the developed image-based material decomposition method accurately estimated 4 basis material density (3 without k-edge and 1 with in the range of the simulated energy bins) even without any prior information about spectra and detector response. This method is applicable to mixtures of solutions and dissolvable materials, where volume conservation assumptions do not apply. CHM receives research support from NIH and Siemens Healthcare.

Multi-energy image sequence fusion based on variable energy X-ray imaging

For complicated structural components characterized by wide X-ray attenuation ranges, the conventional fixed-energy imaging mode cannot obtain all structural information using a single tube voltage. This limitation results in information shortage, because the effective thickness of components along the orientation of the X-ray penetration exceeds the limit of the dynamic range of the X-ray imaging system. To solve this problem, multi-energy image sequence fusion technology has been advanced. In this new method, the tube voltage is adjusted several times by matching the voltage and the effective thickness to obtain all the effective local information on an object. Then, the subset sequences in the multi-energy image sequence are extracted based on the recursive template, and that are fused to reconstruct the full projection information based on linear weighting. An accompanying experiment demonstrates that the new technology can extend the dynamic range of X-ray imaging and provide a complete representation of the internal structure of complicated structural components.

STXM Characterization of PEM Fuel Cell Catalyst Layers

The synchrotron based scanning transmission soft X-ray microscopy (STXM) utilizing near edge X-ray absorption fine structure (NEXAFS) is a novel technique for characterization of membrane electrode assembly (MEA) components in proton exchange membrane fuel cells (PEMFC). Component maps for carbon support, Pt and ionomer in the cathode catalyst layer and perfluorosulphonic acid (PFSA) membrane including the reinforced layer have been obtained from full C 1s multi-energy image sequences and maps from images at four energies in the C 1s and F 1s regions. The ability of this technique to distinguish both structural and chemical differences in cathode catalyst layers from two significantly different designs has been demonstrated. The ionomer distribution in the cathode catalyst layer of an advanced catalyst coated membrane (CCM) structure is compared to that of the cathode in a gas diffusion electrode (GDE).

Quantitative material characterization from multi‐energy photon counting CT

To quantify the concentration of soft-tissue components of water, fat, and calcium through the decomposition of the x-ray spectral signatures in multi-energy CT images. Decomposition of dual-energy and multi-energy x-ray data into basis materials can be performed in the projection domain, image domain, or during image reconstruction. In this work, the authors present methodology for the decomposition of multi-energy x-ray data in the image domain for the application of soft-tissue characterization. To demonstrate proof-of-principle, the authors apply several previously proposed methods and a novel content-aware method to multi-energy images acquired with a prototype photon counting CT system. Data from phantom and ex vivo specimens are evaluated. The number and type of materials in a region can be limited based on a priori knowledge or classification strategies. The proposed difference classifier successfully classified the image into air only, water+fat, water+fat+iodine, and water+calcium regions. Then, the content-aware material decomposition based on weighted least-square optimization generated quantitative maps of concentration. Bias in the estimation of the concentration of water and oil components in a phantom study was <0.10 ± 0.15 g/cc on average. Decomposition of ex vivo carotid endarterectomy specimens suggests the presence of water, lipid, and calcium deposits in the plaque walls. Initial application of the proposed methodology suggests that it can decompose multi-energy CT images into quantitative maps of water, adipose, iodine, and calcium concentrations.

Biological Performance of a Size-Fractionated Core-Shell Tantalum Oxide Nanoparticle X-Ray Contrast Agent

Metal-containing nanoparticles show great promise as x-ray contrast media and could enable reduced radiation dose, increased contrast, and the visualization of smaller anatomic features. In this study, we report progress toward these goals using a size-fractionated core-shell tantalum oxide nanoparticle contrast agent. A core-shell tantalum oxide nanoparticle contrast agent was synthesized and size fractionated for preclinical investigation of biodistribution, blood half-life, organ retention, and histopathology. Fractionated agent was injected at anticipated clinical dose and at 3 times the anticipated clinical dose to evaluate biological performance. Computed tomography (CT) imaging studies were also performed to evaluate short-term clearance kinetics and new imaging applications. Improved control of 2-diethylphosphatoethylsilane-TaO nanoparticle size resulted in significantly reduced retention of injected tantalum. In vivo and in vitro CT imaging studies demonstrated short-term biodistribution differences in the kidney between small-molecule iodinated contrast media and fractionated 2-diethylphosphatoethylsilane-TaO, as well as preliminary data about new "Ta-only" imaging applications using multienergy CT image acquisition. Size-fractionated core-shell tantalum oxide nanoparticles with a well-defined particle size distribution have several key features required of clinically viable vascular imaging compounds and may be used in developing multienergy CT imaging applications.

Multi-energy CT based on a prior rank, intensity and sparsity model (PRISM)

We propose a compressive sensing approach for multi-energy computed tomography (CT), namely the prior rank, intensity and sparsity model (PRISM). To further compress the multi-energy image for allowing the reconstruction with fewer CT data and less radiation dose, the PRISM models a multi-energy image as the superposition of a low-rank matrix and a sparse matrix (with row dimension in space and column dimension in energy), where the low-rank matrix corresponds to the stationary background over energy that has a low matrix rank, and the sparse matrix represents the rest of distinct spectral features that are often sparse. Distinct from previous methods, the PRISM utilizes the generalized rank, e.g., the matrix rank of tight-frame transform of a multi-energy image, which offers a way to characterize the multi-level and multi-filtered image coherence across the energy spectrum. Besides, the energy-dependent intensity information can be incorporated into the PRISM in terms of the spectral curves for base materials, with which the restoration of the multi-energy image becomes the reconstruction of the energy-independent material composition matrix. In other words, the PRISM utilizes prior knowledge on the generalized rank and sparsity of a multi-energy image, and intensity/spectral characteristics of base materials. Furthermore, we develop an accurate and fast split Bregman method for the PRISM and demonstrate the superior performance of the PRISM relative to several competing methods in simulations.

Multi-energy image stack fusion in computed tomography

In the past years computed tomography (CT) was adopted to manufacturing metrology rapidly. Due to the main benefit that lies in the volumetric model generated for each measurement object, many application possibilities from non-destructive testing to coordinate measurements of inner and outer features are possible in much shorter time without destroying the sample. However, some limitations apply to the application of CT. First, it is not possible to measure objects of high aspect ratios (ratio width to thickness) as the changing material thickness influences the remaining intensity after the radiation passed through the workpiece. Secondly, parts consisting of different absorbing materials cannot be measured as only the high absorbing components are acquired with a good signal-to-noise ratio, i.e. with high accuracy and low measurement uncertainty. As high radiation energies are necessary to assure remaining intensities greater than zero in the radiographs, low absorbing components are outshone and not visible after reconstruction. The fusion of multi-energy stacks has been developed to deal with these issues. The approach and first measurement results are shown here.

Apparent image formation by Compton-scattered photons in gamma-ray imaging

The solution to the scatter problem remains a major technical challenge in gamma ray imaging. This paper presents a modelization of apparent images formed by scattered photons at various energies. It may be used either to generate a series of realistic multi-energy images or be included in an image enhancement procedure. The application to SPECT is discussed.