This PDF file contains the front matter associated with SPIE Proceedings Volume 7622, including the Title Page, Copyright information, Table of Contents, and the Conference Committee listing.

Radiation therapy has been on the march toward highly conformal and precise placement of radiation dose within the human body. Development of image-guidance systems based upon x-ray and ultrasound have advanced our capabilities to the level of a few millimeters of uncertainty in many sites. However, there are a subset of existing applications, and potentially many new applications that could be improved upon or pursued if greater targeting accuracy could be achieved. This is motivating the creation of magnetic resonance image-guided radiation treatment units (MRgRT) that either operate simultaneously or provide pre-treatment MR-based targeting capabilities. In this review, the motivation for these pursuits, an overview of recent developments, and a commentary on the challenges they face is presented.

Megavoltage cone-beam computed tomography (MV CBCT) using an electronic portal imaging device (EPID) is a highly promising technique for providing valuable volumetric information for image guidance in radiotherapy. However, active matrix flat-panel imagers (AMFPIs), which are the established gold standard in portal imaging, require a relatively large dose to create images that are clinically useful. This is a consequence of the inefficiency of the phosphor screens employed in conventional MV AMFPIs, which utilize only ~2% of the incident radiation at 6 MV. Fortunately, the incorporation of thick, segmented scintillators can significantly improve the performance of MV AMFPIs, leading to improved image quality for projection imaging at extremely low dose. It is therefore of interest to explore the performance of such thick scintillators for MV CBCT toward the goal of soft-tissue contrast visualization. In this study, prototype AMFPIs incorporating segmented scintillators based on CsI:Tl and BGO crystals with thicknesses ranging from ~11 to 25 mm have been constructed and evaluated. Each prototype incorporates a detector consisting of a matrix of 120 × 60 scintillator elements separated by reflective septal walls, with an element-to-element pitch of 1.016 mm, coupled to an overlying ~1 mm thick Cu plate. The prototype AMFPIs were incorporated into a bench-top CBCT system, allowing the acquisition of tomographic images of a contrast phantom using a 6 MV radiotherapy photon beam. The phantom consists of a water-equivalent (solid water) cylinder, embedded with tissue-equivalent inserts having electron densities, relative to water, varying from ~0.43 to ~1.47. Reconstructed images of the phantom were obtained down to the lowest available dose (one beam pulse per projection), corresponding to a total scan dose of ~4 cGy using 180 projections. In this article, reconstructed images, contrast, noise and contrast-to-noise ratio for the tissue-equivalent objects using the various prototype AMFPIs are examined as a function of dose and compared to those of a conventional MV AMFPI.

Currently, the patient dose of proton therapy and treatment planning are based on the x-ray photon CT (computed tomography) data. However, material interactions between protons and photons are different. Proton tomograms are useful to make with accuracy the dose calculations for planning and positioning of patients. For applying CT techniques, many projection images during at least half rotation are need. It makes extremely high patient dose. We apply the proton tomosynthesis, which is the limited angle tomography using proton-beams. The proton tomosynthesis may provide more accuracy of dose calculation and verifications with less patient dose. We describe the concept of the proton tomosynthesis and demonstrate its performance with GEANT4 simulation and experiments. We confirmed the feasibility of proton digital tomosynthesis through comparative analysis with the x-ray photon methods. This study is useful for proton therapy planning and verification.

We have developed a realistic three-dimensional breast lesion phantom that can be computationally embedded in physically-acquired background images of normal breast tissue. In order to develop new imaging techniques aimed at the detection and diagnosis of breast lesions, a large number of lesions with varying physical characteristics must be tested, especially if physical characteristics must be correlated with observed image features. The new tool presented here, which incorporates three-dimensional tumor features, is potentially useful for testing imaging techniques such as CT, tomosynthesis, and phase-sensitive X-ray imaging, as these require three-dimensional tissue models. The simulated lesions improve significantly upon current methods, which lack the complexity and physical attributes of real tumors, by incorporating a stochastic Gaussian random sphere model to simulate the central tumor mass and calcifications, and an iterative fractal branching algorithm to model the complex spicula present in many tumors. Results show that userdefined lesions with realistic features can be computationally embedded in mammographic background images and that a wide range of physical properties can be modeled.

Analysis of complex imaging tasks requires a phantom that simulates the patient anatomy. We have developed a technique to fabricate 3D physical anthropomorphic breast phantoms for image quality assessment of 2D and 3D breast x-ray imaging systems. The phantom design is based on an existing computer model that can generate breast voxel phantoms of varying size, shape, glandularity, and internal composition. The physical phantom is produced in two steps. First, the computer model of the glandular tissue, skin and Coopers' ligaments is separated into sections. These sections are fabricated by high-resolution rapid prototype printing using a single tissue equivalent material. The adipose tissue regions in the sections are filled using an epoxy-based resin combined with phenolic microspheres. The phantom sections are then stacked. The phantom is provided with an extra section modified to include iodine-enhanced masses. We fabricated a prototype phantom corresponding to a 450 ml breast with 45% dense tissue deformed to represent a 5 cm compressed thickness. The rapid prototype and epoxy based resin phantom materials attenuate x rays similar to 50% glandular tissue and 100% adipose tissue, respectively. The iodinated masses are between 4.0 and 9.6 mm thick and contain 2.5 mg/ml and 5 mg/ml iodine. Digital mammography and digital breast tomosynthesis images of the phantom are qualitatively similar in appearance to clinical images. In summary, a method to fabricate a 3D physical anthropomorphic breast phantom has been developed with known ground truth in the form of a companion voxel phantom. This combined system of physical and computational phantoms allows for both qualitative and quantitative image quality assessment.

Medical imaging equipment is routinely characterised and tested using tissue equivalent phantoms. Combined x-ray and optical mammography could provide increased screening specificity over either system alone. The ongoing evaluation of this approach depends upon the development of phantoms with simultaneously breast tissue equivalent optical and x-ray properties. Furthermore deformation models used in the registration of optical and x-ray images, which are acquired at differing levels of breast compression, require validation through phantoms which are also mechanically tissue equivalent. As well as static imaging, dynamic optical imaging of blood flow whilst breast compression is applied has been proposed as a method of enhancing screening specificity. The effect of changes in blood flow and volume on optical tomography still need to be established. A novel phantom material created by freezing and thawing a solution of polyvinyl alcohol (PVAL) in ethanol to create a solid yet elastically compressible gel is described. These gels have x-ray attenuation coefficients equivalent to those of breast tissues whilst their optical and mechanical properties are readily modified. Titanium dioxide is added to the optically non-scattering and colourless gels to obtain the transport scattering coefficient required. Cancerous tissues are often many times stiffer than healthy. Similar differences in stiffness are achieved between gels by varying PVAL concentration. The first x-ray and optical images of an anthropomorphically shaped breast phantom made from this gel are presented. This contains a lesion filled with blood equivalent dye whose volume changes upon compression of the phantom.

With the injection of iodine, Contrast Enhanced Digital Mammography (CEDM) provides functional information about breast tumour angiogenesis that can potentially help in cancer diagnosis. In order to generate iodine images in which the gray level is proportional to the iodine thickness, temporal and dual-energy approaches have already been considered. The dual-energy method offers the advantage of less patient motion artifacts and better comfort during the exam. However, this approach requires knowledge of the breast thickness at each pixel. Generally, as compression is applied, the breast thickness at each pixel is taken as the compression thickness. Nevertheless, in the breast border region, this assumption is not correct anymore and this causes inaccuracies in the iodine image. Triple-Energy CEDM could overcome these limitations by providing supplemental information in the form of a third image acquired with a different spectrum than the other two. This precludes the need of a priori knowledge of the breast thickness. Moreover, with Triple-Energy CEDM, breast thickness and glandularity maps could potentially be derived. In this study, we first focused on the method to recombine the three images in order to generate the iodine image, analyzing the performance of either quadratic, cubic or conic recombination functions. Then, we studied the optimal acquisition spectra in order to maximize the iodine SDNR in the recombined image for a given target total glandular dose. The concept of Triple-Energy CEDM was validated on simulated textured images and poly-energetic images acquired with a conventional X-ray mammography tube.

With advances in 3D in vivo imaging technology, non-invasive procedures can be used to characterize tissues to identify tumors and monitor changes over time. Using a dedicated breast CT system with a quasi-monochromatic cone-beam x-ray source and flat-panel digital detector, this study was performed in an effort to directly characterize different materials in vivo based on their absolute attenuation coefficients. CT acquisitions were first acquired using a multi-material rod phantom with acrylic, delrin, polyethylene, fat-equivalent, and glandular-equivalent plastic rods, and also with a human cadaver breast. Projections were collected with and without a beam stop array for scatter correction. For each projection, the 2D scatter was estimated with cubic spline interpolation of the average values behind the shadow of each beam stop overlapping the object. Scatter-corrected projections were subsequently calculated by subtracting the scatter images containing only the region of the object from corresponding projections (consisting of primary and scatter x-rays) without the beam stop array. Iterative OSTR was used to reconstruct the data and estimate the non-uniform attenuation distribution. Preliminary results show that with reduced beam hardening from the x-ray beam, scatter correction further reduces the cupping artifact, improves image contrast, and yields attenuation coefficients < 8% of narrow-beam values of the known materials (range 1.2 - 7.8%). Peaks in the histogram showed clear separation between the different material attenuation coefficients. These findings indicate that minimizing beam hardening and applying scatter correction make it practical to directly characterize different tissues in vivo using absolute attenuation coefficients.

Breast density is positively linked to the risk of developing breast cancer. We have developed a semi-automated, stepwedge-based method that has been applied to the mammograms of 1,289 women in the UK breast screening programme to measure breast density by volume and area. 116 images were analysed by three independent operators to assess inter-observer variability; 24 of these were analysed on 10 separate occasions by the same operator to determine intra-observer variability. 168 separate images were analysed using the stepwedge method and by two radiologists who independently estimated percentage breast density by area. There was little intra-observer variability in the stepwedge method (average coefficients of variation 3.49% - 5.73%). There were significant differences in the volumes of glandular tissue obtained by the three operators. This was attributed to variations in the operators' definition of the breast edge. For fatty and dense breasts, there was good correlation between breast density assessed by the stepwedge method and the radiologists. This was also observed between radiologists, despite significant inter-observer variation. Based on analysis of thresholds used in the stepwedge method, radiologists' definition of a dense pixel is one in which the percentage of glandular tissue is between 10 and 20% of the total thickness of tissue.

A new generation of digital breast tomosynthesis system has been designed and is commercially available outside the US. The system has both a 2D mode and a 3D mode to do either conventional mammography or tomosynthesis. Uniquely, it also has a fusion mode that allows both 3D and 2D images to be acquired under the same breast compression, which results in co-registered images from the two modalities. The aim of this paper is to present a technical description on the design and performance of the new system, including system details such as filter options, doses, AEC operation, 2D and 3D images co-registration and display, and the selenium detector performance. We have carried out both physical and clinical studies to evaluate the system. In this paper the focus will be mainly on technical performance results.

Conventional mammography is largely limited by superimposed anatomy which is alleviated by breast tomosynthesis and CT. Limited acquisition in tomosynthesis can result in significant out of plane artifacts while large angular acquisition span in CT can limit the imaging coverage of the chest wall near the breast. We propose a new breast imaging modality, wide-angle breast tomosynthesis (WBT), aimed to provide a practical compromise between 3D sampling and chest-wall coverage. This study compares lesion detection between conventional digital breast tomosynthesis, WBT, and breast CT (44°, 99°, and 198° total angle range, respectively) under equal patient dose conditions. A Monte Carlo (MC) code based on the Penelope package modeled a virtual flat-panel breast tomosynthesis system. The modalities were simulated at four breast compression levels. Glandular dose to the breast was estimated and the radiation flux was subsequently adjusted to achieve a constant mean glandular dose level of 1.5 mGy, independent of the breast thickness and acquisition geometry. Reconstructed volumes were generated using iterative reconstruction methods. Lesion detectability was estimated using contrast-to-noise-ratio. Results showed improved detection with increased angular span and compression. Evaluations also showed improved performance of WBT over DBT at lower compression levels, therefore highlighting potential for reduced breast compression when using a larger acquisition angle.

We are investigating factors affecting the detection of microcalcifications in digital breast tomography (DBT). In this study, we analyzed the effects of projection-view (PV) distribution on spatial blurring of calcifications on the tomosynthesized slices (X-Y plane) and along the depth (Z) direction. DBT scans of a breast phantom with simulated microcalcifications were acquired with a GE prototype system at 21 angles in 3° increments over a ±30° range. Six subsets of 11 PVs were selected from the full set to simulate DBT of different angular ranges and angular increments. SART was applied to each subset to reconstruct the DBT slices. The FWHMs of the line profiles of calcifications within their in-focus DBT slices and FWHMs of the inter-plane artifact spread function (ASF) in the Z-direction for the different PV distributions were compared. The results indicate that DBT acquired with a large angular range or a reasonable number of PVs at large angles yield superior ASF with smaller FWHM in the Z-direction. PV distributions with a narrow angular range have stronger inter-plane artifacts. In the X-Y focal planes, the effect of PV distributions on spatial blurring depends on the directions. The normalized line profiles of the calcifications reconstructed with the different PV distributions are similar in the X-direction. The differences in the FWHMs between the different PV distributions are less than half a pixel. In the Y-(x-ray tube motion) direction, the normalized line profiles of the calcifications reconstructed with DBT acquired with a narrow angular range or a reasonable number of PVs at small angles have less blurring in terms of smaller FWHMs of the line profiles. PV distributions with a wide angular range have stronger in-plane artifacts in the Y-direction. Further study is underway to compare different reconstruction techniques and parameters. The information will be useful for optimization of DBT for detection of microcalcifications.

Digital Breast Tomosynthesis (DBT) suffers from incomplete data and poor quantum statistics limited by the total dose absorbed in the breast. Hence, statistical reconstruction assuming the photon statistics to follow a Poisson distribution may have some advantages. This study investigates state-of-art iterative maximum likelihood (ML) statistical reconstruction algorithms for DBT and compares the results with simple backprojection (BP), filtered backprojection (FBP), and iFBP (FBP with filter derived from iterative reconstruction). The gradient-ascent and convex optimization variants of the transmission ML algorithm are evaluated with phantom and clinical data. Convergence speed is very similar for both iterative statistical algorithms and after approximately 5 iterations all significant details are well displayed, although we notice increasing noise. We found empirically that a relaxation factor between 0.25 and 0.5 provides the optimal trade-off between noise and contrast. The ML-convex algorithm gives smoother results than the ML-gradient algorithm. The low-contrast CNR of the ML algorithms is between CNR for simple backprojection (highest) and FBP (lowest). Spatial resolution of iterative statistical and iFBP algorithms is similar to that of FBP but the quantitative density representation better resembles conventional mammograms. The iFBP algorithm provides the benefits of statistical iterative reconstruction techniques and requires much shorter computation time.

A digital breast tomosynthesis (DBT) reconstruction algorithm has been optimized using an anthropomorphic software breast phantom. The algorithm was optimized in terms of preserving the x-ray attenuation coefficients of the simulated tissues. The appearance of the reconstructed images is controlled in the algorithm using three input parameters related to the reconstruction filter. We varied the input parameters to maximally preserve the attenuation information. The primary interest was to identify and to distinguish between adipose and non-adipose (dense) tissues. To that end, a software voxel phantom was used which included two distinct attenuation values of simulated breast tissues. The phantom allows for great flexibility in simulating breasts of various size, glandularity, and internal composition. Distinguishing between fatty and dense tissues was treated as a binary decision task quantified using ROC analysis. We defined the reconstruction geometry to enable voxel-to-voxel comparison between the original and reconstructed volumes. Separate histograms of the reconstructed pixels corresponding to simulated adipose and non-adipose tissues were computed. ROC curves were generated by varying the reconstructed intensity threshold; pixels above the threshold were classified as dense tissue. The input parameter space was searched to maximize the area under the ROC curve. The reconstructed phantom images optimized in this manner better preserve the tissue x-ray attenuation properties; concordant results are seen in clinical images. Use of the software phantom was successful and practical in this task-based optimization, providing ground truth information about the simulated tissues and providing flexibility in defining anatomical properties.

As a collaborative effort between scientists affiliated with the American Association of Physicists in Medicine (AAPM) and the European reference center for breast cancer screening and diagnosis (EUREF), the Working Group on Phantoms for Breast Imaging (WGPBI) aims to develop phantoms and evaluation techniques for 2D & 3D breast imaging modalities. In the first phase of this collaboration, this project aimed to develop a phantom and associated procedure for constancy testing of digital breast tomosynthesis (DBT) systems. The procedure involves daily and weekly components. The daily evaluation is performed on a simple, homogenous PMMA plate of 4 cm thickness. For the weekly part, a new phantom has been designed consisting of a 45 mm thick homogeneous slab of PMMA with a set of spherical and rectangular inserts at specific 3D positions, and a thin wire positioned at a small angle to the plane of the detector. Quality control parameters are extracted from both projection images (if available) and reconstructed planes. The homogeneous phantom for daily QC allows a trend analysis of homogeneity and the assessment of detector artifacts. With the proposed phantom concept for weekly QC, the stability of the following parameters can be evaluated: the propagation and correlation of the noise in plane and across the reconstructed tomographic planes, lag, signal difference to noise ratio (SDNR) and signal to noise ratio (SNR), the geometry and the motion, effective thickness of the reconstructed planes, homogeneity, distance accuracy, frequency dependent SNR, and artifacts. Analysis of the DICOM header provides information on the stability of the automatic exposure control (AEC), exposure settings, and several system parameters. In an on-going study, the proposed strategy is being applied to five tomosynthesis systems both in Europe and in the US. In this paper we report on the specifics of the phantom, the QC procedure, the practicalities of remote data analysis, and the results of the initial trial.

Digital breast tomosynthesis is a new technique to improve the early detection of breast cancer by providing threedimensional reconstruction volume of the object with limited-angle projection images. This paper investigated the image reconstruction with a standard biopsy training breast phantom using a novel multi-beam X-ray sources breast tomosynthesis system. Carbon nanotube technology based X-ray tubes were lined up along a parallel-imaging geometry to decrease the motion blur. Five representative reconstruction algorithms, including back projection (BP), filtered back projection (FBP), matrix inversion tomosynthesis (MITS), maximum likelihood expectation maximization (MLEM) and simultaneous algebraic reconstruction technique (SART), were investigated to evaluate the image reconstruction of the tomosynthesis system. Reconstructed images of the masses and micro-calcification clusters embedded in the phantom were studied. The evaluated multi-beam X-ray breast tomosynthesis system is able to generate three-dimensional information of the breast phantom with clearly-identified regions of the masses and calcifications. Future study will be done soon to further improve the imaging parameters' measurement and reconstruction.

In this work, analytical models of the optical transfer function (OTF), noise power spectra (NPS), and detective quantum efficiency (DQE) are developed for two types of digital x-ray detectors. The two detector types are (1) energy integrating (EI), for which the point spread function (PSF) is interpreted as a weighting function for counting x-rays, and (2) photon counting (PC), for which the PSF is treated as a probability. The OTF is the Fourier transform of the PSF. The two detector types, having the same PSF, possess an equivalent OTF. NPS is the discrete space Fourier transform (DSFT) of the autocovariance of signal intensity. From first principles, it is shown that while covariance is equivalent for both detector types, variance is not. As a consequence, provided the two detector types have equivalent PSFs, a difference in NPS exists such that NPS_PC ≥ NPS_EI and hence DQE_PC ≤ DQE_EI. The necessary and sufficient condition for equality is that the PSF is either zero or unity everywhere. A PSF modeled as the convolution of a Lorentzian with a rect function is analyzed in order to illustrate the differences in NPS and DQE. The Lorentzian models the blurring of the xray converter, while the rect function reflects the sampling of the detector. The NPS difference between the two detector types is shown to increase with increasing PSF width. In conclusion, this work develops analytical models of OTF, NPS, and DQE for energy integrating and photon counting digital x-ray detectors.

The intrinsic quality of the detection system of 7 different digital mammography units (5 direct radiography DR; 2 computed radiography CR), expressed by DQE, has been compared with their image quality/dose performances in clinical use. DQE measurements followed IEC 62220-1-2 using a tungsten test object for MTF determination. For image quality assessment two different methods have been applied: 1) measurement of contrast to noise ratio (CNR) according to the European guidelines and 2) contrast-detail (CD) evaluation. The latter was carried out with the phantom CDMAM ver. 3.4 and the commercial software CDMAM Analyser ver. 1.1 (both Artinis) for automated image analysis. The overall image quality index IQF_inv proposed by the software has been validated. Correspondence between the two methods has been shown figuring out a linear correlation between CNR and IQF_inv. All systems were optimized with respect to image quality and average glandular dose (AGD) within the constraints of automatic exposure control (AEC). For each equipment, a good image quality level was defined by means of CD analysis, and the corresponding CNR value considered as target value. The goal was to achieve for different PMMA-phantom thicknesses constant image quality, that means the CNR target value, at minimum dose. All DR systems exhibited higher DQE and significantly better image quality compared to CR systems. Generally switching, where available, to a target/filter combination with an x-ray spectrum of higher mean energy permitted dose savings at equal image quality. However, several systems did not allow to modify the AEC in order to apply optimal radiographic technique in clinical use. The best ratio image quality/dose was achieved by a unit with a-Se detector and W anode only recently available on the market.

The MTF, NNPS, and DQE are standard linear system metrics used to characterize intrinsic detector performance. To evaluate total system performance for actual clinical conditions, generalized linear system metrics (GMTF, GNNPS and GDQE) that include the effect of the focal spot distribution, scattered radiation, and geometric unsharpness are more meaningful and appropriate. In this study, a two-dimensional (2D) generalized linear system analysis was carried out for a standard flat panel detector (FPD) (194-micron pixel pitch and 600-micron thick CsI) and a newly-developed, high-resolution, micro-angiographic fluoroscope (MAF) (35-micron pixel pitch and 300- micron thick CsI). Realistic clinical parameters and x-ray spectra were used. The 2D detector MTFs were calculated using the new Noise Response method and slanted edge method and 2D focal spot distribution measurements were done using a pin-hole assembly. The scatter fraction, generated for a uniform head equivalent phantom, was measured and the scatter MTF was simulated with a theoretical model. Different magnifications and scatter fractions were used to estimate the 2D GMTF, GNNPS and GDQE for both detectors. Results show spatial non-isotropy for the 2D generalized metrics which provide a quantitative description of the performance of the complete imaging system for both detectors. This generalized analysis demonstrated that the MAF and FPD have similar capabilities at lower spatial frequencies, but that the MAF has superior performance over the FPD at higher frequencies even when considering focal spot blurring and scatter. This 2D generalized performance analysis is a valuable tool to evaluate total system capabilities and to enable optimized design for specific imaging tasks.

Effective detective quantum efficiency (eDQE) has recently been developed and introduced. The purpose of this study was to evaluate and optimize exposure conditions for chest imaging with digital radiography (DR) using eDQE. The exposure factors that we considered were tube potential and anti-scatter moving grid. The entrance air kerma was adjusted for each tube potential to yield an effective dose of 34 μSv. The scatter fraction (SF), transmission fraction (TF), effective modulation transfer function (eMTF), and effective normalized noise power spectrum (eNNPS) were measured using a phantom that simulates the attenuation and scatter properties of the human chest. Our results suggest that eMTFs are independent of tube potential regardless of whether or not a grid is used. The eNNPSs were greater with the use of an anti-scatter grid than without, while the eDQE was largest at lower tube potential when there was no grid. Our results indicate that the use of low tube potential without an anti-scatter grid is the most appropriate exposure condition for DR in chest imaging. Further research is needed to optimize measurement conditions for other types of imaging studies.

Digital radiography has gained popularity in many areas of clinical practice. This transition brings interest in advancing the methodologies for image quality characterization. However, as the methodologies for such characterizations have not been standardized, the results of these studies cannot be directly compared. The primary objective of this study was to standardize methodologies for image quality characterization. The secondary objective was to evaluate affected factors to Modulation transfer function (MTF), noise power spectrum (NPS), and detective quantum efficiency (DQE) according to image processing algorithm. Image performance parameters such as MTF, NPS, and DQE were evaluated using the international electro-technical commission (IEC 62220-1)-defined RQA5 radiographic techniques. Computed radiography (CR) images of hand posterior-anterior (PA) for measuring signal to noise ratio (SNR), slit image for measuring MTF, white image for measuring NPS were obtained and various Multi-Scale Image Contrast Amplification (MUSICA) parameters were applied to each of acquired images. In results, all of modified images were considerably influence on evaluating SNR, MTF, NPS, and DQE. Modified images by the post-processing had higher DQE than the MUSICA=0 image. This suggests that MUSICA values, as a post-processing, have an affect on the image when it is evaluating for image quality. In conclusion, the control parameters of image processing could be accounted for evaluating characterization of image quality in same way. The results of this study could be guided as a baseline to evaluate imaging systems and their imaging characteristics by measuring MTF, NPS, and DQE.

The basic principles of x-ray image formation in radiography have remained essentially unchanged since R¨ontgen first discovered x-rays over a hundred years ago. The conventional approach relies on x-ray absorption as the sole source of contrast and draws exclusively on ray or geometrical optics to describe and interpret image formation. This approach ignores another, potentially more useful source of contrast, namely phase and scattering information. Phase-contrast imaging techniques, which can be understood using wave optics rather than ray optics, offer ways to augment or complement standard absorption contrast by incorporating phase information. The recent development of grating based phase- and darkfield-contrast imaging with x-rays1 pawed the way for many potential applications to medical imaging and structure determination in material science. Here we present our recent contributions to the field of interferometric phase-contrast and dark-field x-ray imaging. We introduce a new material dependent scattering parameter, the Linear Diffusion Coefficient, and a quantitative mathematical formalism to extend the dark-field x-ray images into three dimensions by tomographic reconstruction. Further, the results of two experiments that illustrate the potential of dark-field imaging for computed tomography are shown.

Compared to single energy CT, which provides information only about the x-ray linear attenuation coefficients, dual energy CT is able to obtain the electron density and effective atomic number for different materials in a quantitative way. In this study, as an alternative to dual energy CT, a novel quantitative imaging method based on phase contrast CT is described. Rather than requiring two scans with different x-ray photon energies, diffraction grating-based phase contrast CT is capable of reconstructing images of both the linear attenuation and refractive index decrement from a single scan. From the two images, quantitative information of both the electron density and effective atomic number can be extracted. Experimental results demonstrate that: (1) electron density can be accurately determined from refractive index decrement through a linear relationship; and (2) effective atomic number can be explicitly derived from the ratio of linear attenuation to refractive index decrement, using a simple function, i.e., a power function plus a constant. The presented method will shed insight into the field of material separation and find its use in medical and non-medical applications.

Dark-field x-ray projection imaging and dark-field neutron computed tomography have both recently been demonstrated. Such techniques provide insight into the small-angle scattering properties of the image objects. In this work, the dark-field x-ray imaging method is extended to x-ray computed tomography in order to provide unique and complementary information to the previously reported phase contrast and absorption contrast CT images. Dark-field reconstructions are presented and compared to these two other contrast mechanisms, all three of which can be obtained from a single acquisition using an x-ray grating interferometer setup. Objects which provide little absorption contrast, but have a significant small-angle scattering component, are better visualized in a dark-field CT reconstruction.

We present a simulation framework for X-ray phase-contrast computed tomography imaging (PCTI) inheriting the wave- as well as the particle-behavior of photons. The developed tool includes the modeling of a partially coherent X-ray source, the propagation of the X-ray photons through samples, and the interfering properties of photons. Hence, the simulation is capable of physically modeling a grating-based interferometric imaging system reported in e.g. Pfeiffer et al.⁵ The information gained comprises the three potentially measurable images, which are the absorption image, the phase image, and the darkfield image. Results on such a setup concerning spatial and temporal coherence will be shown. Samples consisting of elements and structures similar to biological tissue were implemented to demonstrate the applicability on medical imaging. For the purpose of CT-imaging a head-like phantom was simulated and the results show the advantage of PCTI for thick biological objects. The simulation was developed with a modular concept so that the influences of each imaging component can be considered seperately. Thus the grating based interferometry for X-ray phase-contrast imaging can be optimized towards dedicated medical applications using this simulation-tool.

Propagation-based X-ray phase-contrast imaging permits the visualization of tissues that have very similar X-ray absorption properties and may benefit a variety of biomedical imaging applications. Unlike conventional radiographic contrast that is related to the projected absorption properties of tissue, image contrast in phasecontrast radiographs contains contributions from absorption-contrast and phase-contrast. In this work, we develop a general theoretical framework for assessing the contributions of these contrast mechanisms to signal detectability measures in propagation-based X-ray phase-contrast imaging. Specifically, concepts from signal detection theory are utilized to analyze the contributions of phase- and absorption-contrast to an ideal observer figure of merit for a signal-known-exactly/background-known exactly detection task.

The current estimations of risk associated with medical imaging procedures rely on assessing the organ dose via direct measurements or simulation. Each organ dose is assumed to be homogeneous, a representative sample or mean of which is weighted by a corresponding tissue weighting factor provided by ICRP publication 103. The weighted values are summed to provide Effective Dose (ED), the most-widely accepted surrogate for population radiation risk. For individual risk estimation, one may employ Effective Risk (ER), which further incorporates gender- and age-specific risk factors. However, both the tissue-weighting factors (as used by ED) and the risk factors (as used by ER) were derived (mostly from the atomic bomb survivor data) under the assumption of a homogeneous dose distribution within each organ. That assumption is significantly violated in most medical imaging procedures. In chest CT, for example, superficial organs (eg, breasts) demonstrate a heterogeneous distribution while organs on the peripheries of the irradiation field (eg, liver) possess a nearly discontinuous dose profile. Projection radiography and mammography involve an even wider range of organ dose heterogeneity spanning up to two orders of magnitude. As such, mean dose or point measured dose values do not reflect the maximum energy deposited per unit volume of the organ, and therefore, effective dose or effective risk, as commonly computed, can misrepresent irradiation risk. In this paper, we report the magnitude of the dose heterogeneity in both CT and projection x-ray imaging, provide an assessment of its impact on irradiation risk, and explore an alternative model-based approach for risk estimation for imaging techniques involving heterogeneous organ dose distributions.

This work optimized a multi-pinhole collimator for a stationary three-camera Single Photon Emission Computed Tomography (SPECT) system designed for rapid (one-second) dynamic imaging through simulations. Multi-pinhole collimator designs were investigated to increase efficiency and angular sampling while maintaining adequate spatial resolution for small-animal imaging. The study first analytically investigated the tradeoffs between efficiency and spatial resolution as a function of the number of pinholes. An oval arrangement of pinholes was proposed, and the benefits compared to a circular arrangement were quantified through simulations. Finally, collimators with six to nine pinholes were simulated, and the resulting data compared with respect to efficiency, signal-to-noise ratio (SNR), and the angular coverage in Radon space. All simulations used the GATE Monte Carlo toolkit. The results suggest that an oval arrangement of nine pinholes improved the efficiency and SNR by factors of 1.65 and 1.3, respectively, compared to a circular arrangement. A nine-pinhole collimator was found to provide the highest geometric efficiency (~6.35*10⁴cps/mCi) and improved the SNR by a factor of ~1.3 and ~1.1 compared to the six- and eight-pinhole collimators, respectively. Overall, the simulated multi-pinhole system depicted cylindrical objects despite the limited angular sampling and scan time of the one-second, stationary three-camera acquisition.

The goal of the study was to investigate data acquisition strategies and image reconstruction methods for a stationary SPECT insert that can operate inside an MRI scanner with a 12 cm bore diameter for simultaneous SPECT/MRI imaging of small animals. The SPECT insert consists of 3 octagonal rings of 8 MR-compatible CZT detectors per ring surrounding a multi-pinhole (MPH) collimator sleeve. Each pinhole is constructed to project the field-of-view (FOV) to one CZT detector. All 24 pinholes are focused to a cylindrical FOV of 25 mm in diameter and 34 mm in length. The data acquisition strategies we evaluated were optional collimator rotations to improve tomographic sampling; and the image reconstruction methods were iterative ML-EM with and without compensation for the geometric response function (GRF) of the MPH collimator. For this purpose, we developed an analytic simulator that calculates the system matrix with the GRF models of the MPH collimator. The simulator was used to generate projection data of a digital rod phantom with pinhole aperture sizes of 1 mm and 2 mm and with different collimator rotation patterns. Iterative ML-EM reconstruction with and without GRF compensation were used to reconstruct the projection data from the central ring of 8 detectors only, and from all 24 detectors. Our results indicated that without GRF compensation and at the default design of 24 projection views, the reconstructed images had significant artifacts. Accurate GRF compensation substantially improved the reconstructed image resolution and reduced image artifacts. With accurate GRF compensation, useful reconstructed images can be obtained using 24 projection views only. This last finding potentially enables dynamic SPECT (and/or MRI) studies in small animals, one of many possible application areas of the SPECT/MRI system. Further research efforts are warranted including experimentally measuring the system matrix for improved geometrical accuracy, incorporating the co-registered MRI image in SPECT reconstruction, and exploring potential applications of the simultaneous SPECT/MRI SA system including dynamic SPECT studies.

We previously designed a component based 3-D PSF model to obtain a compact yet accurate system matrix for a dedicated human brain PET scanner. In this work, we adapted the model to a small animal PET scanner. Based on the model, we derived the system matrix for back-to-back gamma source in air, fluorine-18 and iodine-124 source in water by Monte Carlo simulation. The characteristics of the PSF model were evaluated and the performance of the newly derived system matrix was assessed by comparing its reconstructed images with the established reconstruction program provided on the animal PET scanner. The new system matrix showed strong PSF dependency on the line-of-response (LOR) incident angle and LOR depth. This confirmed the validity of the two components selected for the model. The effect of positron range on the system matrix was observed by comparing the PSFs of different isotopes. A simulated and an experimental hot-rod phantom study showed that the reconstruction with the proposed system matrix achieved better resolution recovery as compared to the algorithm provided by the manufacturer. Quantitative evaluation also showed better convergence to the expected contrast value at similar noise level. In conclusion, it has been shown that the system matrix derivation method is applicable to the animal PET system studied, suggesting that the method may be used for other PET systems and different isotope applications.

The recent findings of high heterogeneity of human breast tissue and much lower than predicted dielectric contrast between tumors and their host tissue have raised questions about the potential utility of stand-alone microwave breast imaging techniques. Multimodal approaches that employ microwaves together with other imaging techniques seem more promising. This study investigates a CT-microwave combination in which microwave detection makes use of prior information obtained from volumetric CT scans and knowledge of tissue dielectric properties. In particular, a detailed patient-specific tissue distribution is first obtained from a 3D-CT scan of the breast under exam. It is assumed that from this scan a limited suspect region is identified. Then from recent research results on the dielectric properties of breast tissue, complex permittivity (dielectric constant and conductivity) maps of the breast can be constructed under the hypotheses of normal and cancerous tissue in the suspect region. These in turn can be used with electromagnetic (EM) simulation software to generate empirical distributions for the microwave system observations under each hypothesis. Microwave detection is then performed. Instead of trying to recover a complete dielectric image of the breast from the microwave scan, the question of interest in this approach is simply which hypothesis is more consistent with the observed electromagnetic response of the microwave system. A hypothesis testing method based on the likelihood ratio for the empirical distributions and Receiver Operating Characteristic (ROC) optimization is proposed. The results from a simple idealized test case show good potential and invite further study.

Purpose: In the early development of new imaging modalities - such as tomosynthesis and cone-beam CT (CBCT) - an accurate predictive model for imaging performance is particularly valuable in identifying the physical factors that govern image quality and guiding system optimization. In this work, a task-based cascaded systems model for detectability index is proposed that describes not only the signal and noise propagation in the 2D (projection) and 3D (reconstruction) imaging chain but also the influence of background anatomical noise. The extent to which generalized detectability index provides a valid metric for imaging performance was assessed through direct comparison to human observer experiments. Methods: Detectability index (d') was generalized to include anatomical background noise in the same manner as the generalized noise-equivalent quanta (NEQ) proposed by Barrett et al. (Proc. SPIE Med. Imaging, Vol. 1090, 1989). Anatomical background noise was measured from a custom phantom designed to present power-law spectral density comparable to various anatomical sites (e.g., breast and lung). Theoretical calculations of d' as a function of the sourcedetector orbital extent (θ_tot) was obtained from a 3D cascaded systems analysis model for tomosynthesis and cone-beam CT (CBCT). Four model observers were considered in the calculation of d': prewhitening (PW), non-prewhitening (NPW), prewhitening with eye filter and internal noise (PWE), and non-prewhitening with eye filter and internal noise (NPWE). Human observer performance was measured from 9AFC tests for a variety of idealized imaging tasks presented within a clutter phantom. Theoretical results (d') were converted to area under the ROC curve (Az) and compared directly to human observer performance as a function of imaging task and orbital extent. Results: Theoretical results demonstrated reasonable correspondence with human observer response for all tasks across the continuum in θ_tot ranging from low-angle tomosynthesis (θ_tot ~10^o) to CBCT (θ_tot ~180^o). Both theoretical and experimental Az were found to increase with acquisition angle, consistent with increased rejection of out-of-plane clutter for larger tomosynthesis angle. Of the four theoretical model observers considered, the prewhitening models tended to overestimate real observer performance, while the non-prewhitening models demonstrated reasonable agreement. Conclusions: Generalized detectability index was shown to provide a meaningful metric for imaging performance, helping to bridge the gap between real observer performance and prevalent Fourier-based metrics based in first principles of spatial-frequency-dependent NEQ and imaging task.

This study aimed to extend Fourier-based imaging metrics for the modeling of quantitative imaging performance. Breast tomosynthesis was used as a platform for investigating acquisition and processing parameters (e.g., acquisition angle and dose) that can significantly affect 3D signal and noise, and consequently quantitative imaging performance. The detectability index was computed using the modulation transfer function and noise-power spectrum combined with a Fourier description of imaging task. Three imaging tasks were considered: detection, area estimation (in coronal slice), and volume estimation of a 4 mm diameter spherical target. Task functions for size estimation were generated by using measured performance of the maximum-likelihood estimator as training data. The detectability index computed with the size estimation tasks correlated well with precision measurements for area and volume estimation over a fairly broad range of imaging conditions and provided a meaningful figure of merit for quantitative imaging performance. Furthermore, results highlighted that optimal breast tomosynthesis acquisition parameters depend significantly on imaging task. Mass detection was optimal at an acquisition angle of 85° while area and volume estimation for the same mass were optimal at ~100° and 125° acquisition angles, respectively. These findings provide key initial validation that the Fourier-based detectability index extended to estimation tasks can represent a meaningful metric and predictor of quantitative imaging performance.

Spectral imaging is a method in medical x-ray imaging to extract information about the object constituents by the material-specific energy dependence of x-ray attenuation. Contrast-enhanced spectral imaging has been thoroughly investigated, but unenhanced imaging may be more useful because it comes as a bonus to the conventional non-energy-resolved absorption image at screening; there is no additional radiation dose and no need for contrast medium. We have used a previously developed theoretical framework and system model that include quantum and anatomical noise to characterize the performance of a photon-counting spectral mammography system with two energy bins for unenhanced imaging. The theoretical framework was validated with synthesized images. Optimal combination of the energy-resolved images for detecting large unenhanced tumors corresponded closely, but not exactly, to minimization of the anatomical noise, which is commonly referred to as energy subtraction. In that case, an ideal-observer detectability index could be improved close to 50% compared to absorption imaging. Optimization with respect to the signal-to-quantum-noise ratio, commonly referred to as energy weighting, deteriorated detectability. For small microcalcifications or tumors on uniform backgrounds, however, energy subtraction was suboptimal whereas energy weighting provided a minute improvement. The performance was largely independent of beam quality, detector energy resolution, and bin count fraction. It is clear that inclusion of anatomical noise and imaging task in spectral optimization may yield completely different results than an analysis based solely on quantum noise.

In this study, we examine the performance of the simultaneous algebraic reconstruction technique (SART) for digital breast tomosynthesis under variations in key imaging parameters, such as the number of iterations, number of projections, angular range, initial guess, radiation dose, etc. We use a real breast CT volume as a ground truth digital phantom from which to simulate x-ray projections under the various selected conditions. The reconstructed image quality is measured using task-based metrics, namely signal CNR and the AUC of a Channelised Hotelling Observer with Laguerre-Gauss basis functions. The task at hand is a signal-known-exactly (SKE) task, where the objective is to detect a simulated mass inserted into the breast CT volume.

Pixel Signal to Noise Ratio (SNR) is a commonly used clinical metric for evaluating mammography. However, we showed in this paper, the pixel SNR can produce misleading system detectability when image processing is utilized. We developed a simple, reliable and clinically applicable methodology to evaluate mammographic imaging systems using a task SNR that accounts for the imaging system performance in the presence of the patient. We used the Hotelling observer method in spatial frequency domain to calculate the task SNR of small disk test objects embedded in the breast tissue-equivalent series (BRTES) phantom for GE Senographe DS Full Field Digital Mammography (FFDM) system. The results were compared to the calculation of pixel SNR. We calculated the Hotelling observer SNR by estimating the generalized modulation transfer function (GMTF), generalized normalized noise power spectrum (GNNPS) and generalized noise equivalent quanta (GNEQ) in the presence of the breast phantom. The task SNR we calculated increased with the square root of the exposure as expected. Furthermore, we showed that the method is stable under image processing. The task SNR is a more reliable method for evaluating the performance of imaging systems especially under realistic clinical conditions where patient equivalent phantoms or image processing is used.

Quantitative measures of image quality (IQ) are routinely obtained during the evaluation of imaging systems. These measures, however, do not necessarily correlate with the IQ of the actual clinical images, which can also be affected by factors such as patient positioning. No quantitative method currently exists to evaluate clinical IQ. Therefore, we investigated the potential of using computerized image texture analysis to quantitatively assess IQ. Our hypothesis is that image texture features can be used to assess IQ as a measure of the image signal-to-noise ratio (SNR). To test feasibility, the "Rachel" anthropomorphic breast phantom (Model 169, Gammex RMI) was imaged with a Senographe 2000D FFDM system (GE Healthcare) using 220 unique exposure settings (target/filter, kVs, and mAs combinations). The mAs were varied from 10%-300% of that required for an average glandular dose (AGD) of 1.8 mGy. A 2.5cm² retroareolar region of interest (ROI) was segmented from each image. The SNR was computed from the ROIs segmented from images linear with dose (i.e., raw images) after flat-field and off-set correction. Image texture features of skewness, coarseness, contrast, energy, homogeneity, and fractal dimension were computed from the Premium View^TM postprocessed image ROIs. Multiple linear regression demonstrated a strong association between the computed image texture features and SNR (R²=0.92, p≤0.001). When including kV, target and filter as additional predictor variables, a stronger association with SNR was observed (R²=0.95, p≤0.001). The strong associations indicate that computerized image texture analysis can be used to measure image SNR and potentially aid in automating IQ assessment as a component of the clinical workflow. Further work is underway to validate our findings in larger clinical datasets.

We present a Monte Carlo (MC) simulation method for studying the signal formation process in amorphous Selenium (a-Se) imaging detectors for design validation and optimization of direct imaging systems. The assumptions and limitations of the proposed and previous models are examined. The PENELOPE subroutines for MC simulation of radiation transport are used to model incident x-ray photon and secondary electron interactions in the photoconductor. Our simulation model takes into account applied electric field, atomic properties of the photoconductor material, carrier trapping by impurities, and bimolecular recombination between drifting carriers. The particle interaction cross-sections for photons and electrons are generated for Se over the energy range of medical imaging applications. Since inelastic collisions of secondary electrons lead to the creation of electron-hole pairs in the photoconductor, the electron inelastic collision stopping power is compared for PENELOPE's Generalized Oscillator Strength model with the established EEDL and NIST ESTAR databases. Sample simulated particle tracks for photons and electrons in Se are presented, along with the energy deposition map. The PENEASY general-purpose main program is extended with custom transport subroutines to take into account generation and transport of electron-hole pairs in an electromagnetic field. The charge transport routines consider trapping and recombination, and the energy required to create a detectable electron-hole pair can be estimated from simulations. This modular simulation model is designed to model complete image formation.

The imager presented in this paper has a special blocking structure that ensures very low dark current of less than 1 pA/mm² even with a 20 V/μm electric field. Hence the electric field can be increased from the generally applied 10 V/μm to 20V/μm, this reduces the energy required to produce an electron hole (e-h) pair from 60 eV to about 36 eV at the given (19.3 keV mean) mammo energy. Furthermore, with special doping and manufacturing processes this a-Se layer is very stable in the 0-70 C° temperature range as demonstrated by Ogusu et al. [1]. A new 5 cm × 5 cm size TFT array was developed with 50 μm pixel size, specifically for testing the resolution of photoconductor based imagers. The first new imager of this type had a 200 μm thick a-Se layer evaporated onto the array. Its MTF, NPS, and DQE values were evaluated using 28kVp Mo anode x-ray source with a 0.03mm thick Mo and an additional 2 mm thick Al filters. The MTF value is about 40% and 50% in x-and y directions at the Nyquist frequency of 10 lp/mm. The low frequency DQE at 20 V/μm electrical field is ~70% at 151 μGy dose and drops only about 10% when going down to 23 μGy. This new array also has excellent lag properties. The measured first frame image lag at 20 V/μm is less than 1%. Such low lag provides opportunities to use this material not only for mammography but also for breast tomosynthesis applications. Breast phantom images demonstrate that even the smallest 0.13 mm calcifications are clearly visible with this high-resolution imager.

The ghosting recovery techniques and mechanisms in multilayer selenium X-ray detector structures for mammography are experimentally and theoretically investigated. The experiments have been carried out under low positive applied electric field (~1-2V/μm) since a very little ghost can be seen under normal operating applied electric field (10V/μm). A ghost removal technique is investigated by reversing the bias polarity during the natural recovery process. The theoretical model considers accumulated trapped charges and their effects (trap filling, recombination, electric field profile, and electric field dependent electron-hole pair creation), the carrier transport in the blocking layers, and the effects of charge injection from the metal contacts. We consider carrier trapping in both charged and neutral defect states. It has been assumed that the X-ray induced deep trap centers are neutral defects. The time dependent carrier detrapping and structural relaxation (recovery of meta-stable trap centers) are also considered. The sensitivity in a rested sample is recovered mainly by the carrier detrapping, the recombination of the injected carriers with the existing trapped carriers, and the relaxation of the X-ray induced deep trap centers. A faster sensitivity recovery is found by reversing the bias during the natural recovery process. During reverse bias huge number of holes are injected from the metal and recombine with the trapped electrons. This results in faster sensitivity recovery. The electric fields at the metal contacts increase with time at the beginning of the natural ghosting recovery process which leads to the initial increase of the dark current. Later the electric fields at the metal contacts decrease and hence the dark current decays over time during the natural recovery process. The theoretical model shows a very good agreement with the experimental results.

Thick amorphous selenium (a-Se) as an excellent photoconductor has been used in direct conversion X-ray imaging modalities such as mammography. However, due to substantial charge trapping, such detectors experience a long X-ray response time and as a result, suffer from a slow speed of operation. Therefore, its deployment to speed-required applications such as real-time fluoroscopy remains a challenge. In this work, we aim to investigate a lateral a-Se MSM photodetector as an indirect conversion X-ray imager and its utilization in high speed, high energy medical applications. The dark current density of the newly-fabricated detector is below 20 pA/mm2 for a 200 μm×50 μm pixel pitch at electric field strengths ranging from 6 to 12 V/μm. The photoresponsivity reaches up to 2.3A/W towards blue wavelength of 468 nm at an electric field strength of 20 V/μm. Furthermore, the photocurrent has a fast speed of photoresponse, demonstrating rise time, fall time and time constant of 50 μs, 60 μs and 30 μs, respectively. Given that low dark current and high photoresponsivity this detector holds, coupled with fast photoresponse, it is believed that lateral a-Se MSM photodetector is promising for indirect conversion X-ray imager integrated with either CMOS or TFT arrays.

Flat panel selenium detectors (1) have been commercially available since 1998 (2). The MTF of these detectors can approach the theoretical SINC function for the pixel size (3). Detectors can be designed with selenium thickness suitable for absorption of the range of x-ray energy for the modality (4, 5). For higher energy x rays, the thickness of the selenium layer can be increased without greatly degrading the spatial resolution. The non-spreading nature of the signal allows the detector to detect very weak x-ray signal in the vicinity of strong signal. Selenium detectors can therefore be designed to produce very high dynamic range images when needed. However, as a photo-conducting material, selenium also comes with some less than ideal properties. For example, charge trapping, long settling time for with bias electric field, and interface charge injection (6). These adverse properties must be included in detector design for optimal performance in each application. This paper describes a novel method for interfacial charge removal using lateral conductivity of selenium.

A generalized approach to describing transfer of signal and noise (MTF and NPS) through medical imaging systems has been developed over the past several years in which image-forming processes are represented in terms of serial and parallel cascades of amplified point processes. We use the techniques of both cascaded systems analysis and stochastic point process theory to develop fundamental limitations of system performance for single photon counting (SPC) x-ray imaging detectors to assist in the optimal design of new systems. Using this approach the mean signal and signal variance for a simple model of a hypothetical flat-panel x-ray imaging detector are calculated. It is shown that energy imprecision is ultimately determined by the number of secondary quanta collected by the detector. Successful designs will likely have small work-function values and/or high collection efficiencies as well as adaptive binning strategies to collect all secondary quanta liberated for each x-ray interaction. K-escape, responsible for Swank noise in traditional imging detectors, can result in increases of relative imprecision to 50% and will impose a severe limitation on the ability of these detectors to determine x-ray spectra and methods will have to be incorporated into new detector designs to overcome these limitations. With the exception of K-escape effects, Se-based and CsI-based detectors could potentially measure spectra with energy RMS imprecision of 3 - 10%.

A single photon counting Voltage Controlled Oscillator (VCO) based pixel architecture in amorphous silicon (a-Si) technology is reported for large area digital medical imaging. The VCO converts X-ray generated input charge into an output oscillating frequency signal. Experimental results for an in-house fabricated VCO circuit in a-Si technology are presented and external readout circuits to extract the image information from the VCO's frequency output are discussed. These readout circuits can be optimized to reduce the fixed pattern noise and fringing effects in an imaging array containing many such VCO pixels. Noise estimations, stability simulations and measurements for the fabricated VCO are presented. The reported architecture is particularly promising for large area photon counting applications (e.g. low dose fluoroscopy, dental computed tomography (CT)) due to its very low input referred electronic noise, high sensitivity and ease of fabrication in low cost a-Si technology.

The overall aim of this work was to evaluate the potential for improving in vivo small animal microCT through the use of an energy resolved photon-counting detector. To this end, we developed and evaluated a prototype microCT system based on a second-generation photon-counting x-ray detector which simultaneously counted photons with energies above six energy thresholds. First, we developed a threshold tuning procedure to reduce the dependence of detector uniformity and to reduce ring artifacts. Next, we evaluated the system in terms of the contrast-to-noise ratio in different energy windows for different target materials. These differences provided the possibility to weight the data acquired in different windows in order to optimize the contrast-to-noise ratio. We also explored the ability of the system to use data from different energy windows to aid in distinguishing various materials. We found that the energy discrimination capability provided the possibility for improved contrast-to-noise ratios and allowed separation of more than two materials, e.g., bone, soft-tissue and one or more contrast materials having K-absorption edges in the energy ranges of interest.

Recently, a novel CdTe photon counting x-ray detector (PCXD) with energy discrimination capabilities has been developed [1, 2]. When such detectors are operated under a high count rate, however, coincident pulses and tails of pulses distort the recorded energy spectrum. These distortions are called pulse pileup effects. It is essential to compensate for these effects on the recorded energy spectrum in order to take a full advantage of spectral information PCXDs provide. In this study, we have developed a new analytical pulse pileup model which uses a model of the measured pulse shape measured for a PCXD [1, 2]. We validated the model using Monte Carlo simulations of monochromatic and polychromatic spectra compared to predictions from our model. Excellent agreement was found between the recorded spectra obtained by the MC simulations and those calculated by our analytical model.

Spectral X-ray imaging is a promising technique to drastically improve the diagnostic quality of radiography and computed tomography (CT), since it enables material decomposition and/or identification based on the energy dependency of material-specific X-ray attenuation. Unlike the charge-integration based X-ray detectors, photon counting X-ray detectors (PCXDs) can discriminate the energies of incident X-ray photons and thereby multi-energy images can be obtained in single exposure. However, the measured data are not accurate since the spectra of incident X-rays are distorted according to the energy response function (ERF) of a PCXD. Thus ERF should be properly estimated in advance for accurate spectral imaging. This paper presents a simple method for ERF estimation based on a polychromatic X-ray source that is widely used for medical imaging. The method consists of three steps: source spectra measurement, detector spectra reconstruction, and ERF inverse estimation. Real spectra of an X-ray tube are first measured at all kVs by using an X-ray spectrometer. The corresponding detector spectra are obtained by threshold scans. The ERF is then estimated by solving the inverse problem. Simulations are conducted to demonstrate the concept of the proposed method.

We report results from the development of a second-generation CdTe direct-conversion compound-semiconductor x-ray detector for photon-counting clinical CT. The first-generation detector has 512 pixels with a 1 mm pitch and is vertically integrated with the readout. A 32-row multi-slice CT system using first-generation detectors has been used for clinical low-dose CT applications. To provide adequate performance for whole-body diagnostic CT we have designed and fabricated new 0.25 mm² pixels to increase the maximum output to greater than 20 Mcps per mm² while preserving sufficient energy resolution for photon-counting CT. In addition to the need for dynamic range, CT places stringent uniformity and temporal response requirements on the detector. We have measured detector parameters including the dynamic range, energy resolution, noise floor, stability, and temporal response. Temporal response is determined by rapid cycling of the input flux with shutter driven attenuators. Cycling between high and low flux generates reproducible counts, within counting statistics, with a response time less than 1 ms. Stability is determined by measuring uniformity corrected flood images repeatedly over a time interval exceeding whole-body diagnostic CT scan times. Long exposure to uniform flux generates a number of counts which drift in some pixels slightly in excess of counting statistics. These results demonstrate the potential for these detectors to achieve whole-body CT.

Perfusion computed tomography is increasingly being used for stroke and tumor assessment. Using continuous periodic table movement the spatial coverage can be increased beyond the detector width with a new adaptive spiral scanning technique (A4DS). The purpose of this study was to compare perfusion values acquired with the A4DS technique with results from standard dynamic scans at different temporal sampling rates. A biological perfusion phantom (preserved porcine kidney) was scanned with both techniques. In standard mode three scans were performed at adjacent overlapping positions (detector width 38.4 mm) covering the whole phantom. Data were reconstructed with temporal resolutions of 0.5, 1 and 1.5 s. The A4DS scan was performed with a cycle time of 1.5 s and scan ranges of 100 and 148 mm respectively. The phantom was not repositioned between scans in order to assure that identical image slices showed identical phantom slices. Tissue flow was calculated with a deconvolution type algorithm. Regions of interest were drawn in strongly and moderately enhancing areas and around the whole cortex in three slices in the upper, central and lower portion of the phantom. In the flow range of 40 to 100 ml/100ml/min values did not differ by more than 5 ml/100ml/min between any of the scan protocols used. The correlation between the continuous table movement modes and the 0.5 s standard mode was excellent (r²>0.98) indicating that the new mode is well suited for perfusion measurements and allows increasing the coverage by almost a factor of four.

As a widely adopted imaging modality for pre-clinical research, micro-CT is constantly facing the need of providing better temporal as well as spatial resolution for a variety of imaging applications. Faster CT scanning speed is also preferred for higher imaging throughput. We recently proposed a gantry-free multi-beam micro-CT (MBμCT) design which has the potential to overcome some of the intrinsic limitations of current rotating-gantry CT technology. To demonstrate its feasibility, we have constructed a testing system with a multi-beam field emission x-ray (MBFEX) source array with a linear array of 20 individually controllable x-ray emitting pixels. Based on simulations of the electron optics and preliminary experimental measurements the design of the MBFEX source has been further optimized. The newly designed imaging system has been characterized and commissioned following our standard imaging protocol. It has clearly shown improved system stability and enhanced imaging capability. As a result of reduced mechanical rotation during imaging acquisition, we are expecting to achieve higher CT scanning speed without significantly sacrificing imaging quality. This prototype MBμCT system, although still in its early development phase, has been proved to be an ideal testing platform for the proposed gantry-free micro-CT scanner.

This paper summarizes the development of a distributed x-ray source with up to 60kW demonstrated instantaneous power. Component integration and test results are shown for the dispenser cathode electron gun, fast switching controls, high voltage stand-off insulator, brazed anode, and vacuum system. The current multisource prototype has been operated for over 100 hours without failure, and additional testing is needed to discover the limiting component. Example focal spot measurements and x-ray radiographs are included. Lastly, future development opportunities are highlighted.

This paper describes the image quality improvements achieved by developing a new fully physical imaging chain. The key enablers for this imaging chain are a new scatter correction technique and an analytic computation of the beam hardening correction for each detector. The new scatter correction technique uses off-line Monte Carlo simulations to compute a large database of scatter kernels representative of a large variety of patient shapes and an on-line combination of those based on the attenuation profile of the patient in the measured projections. In addition, profiles of scatter originating from the wedge are estimated and subtracted. The beam hardening coefficients are computed using analytic simulations of the full beam path of each individual ray through the scanner. Due to the new approach, scatter and beam hardening are computed from first principles with no further tuning factors, and are thus straight forward to adapt to any patient and scan geometry. Using the new fully physical imaging chain unprecedented image quality was achieved. This is demonstrated with a special scatter phantom. With current image correction techniques this phantom typically shows position dependent inhomogeneity and streak artifacts resulting from the impact of scattered radiation. With the new imaging chain these artifacts are almost completely eliminated, independent of position and scanning mode (kV). Further preliminary patient studies show that in addition to fully guaranteeing an absolute Hounsfield scale in arbitrary imaging conditions, the new technique also strongly sharpens object boundaries such as the edges of the liver.

Radiation-dose awareness and optimization in CT can greatly benefit from a dosereporting system that provides radiation dose and cancer risk estimates specific to each patient and each CT examination. Recently, we reported a method for estimating patientspecific dose from pediatric chest CT. The purpose of this study is to extend that effort to patient-specific risk estimation and to a population of pediatric CT patients. Our study included thirty pediatric CT patients (16 males and 14 females; 0-16 years old), for whom full-body computer models were recently created based on the patients' clinical CT data. Using a validated Monte Carlo program, organ dose received by the thirty patients from a chest scan protocol (LightSpeed VCT, 120 kVp, 1.375 pitch, 40-mm collimation, pediatric body scan field-of-view) was simulated and used to estimate patient-specific effective dose. Risks of cancer incidence were calculated for radiosensitive organs using gender-, age-, and tissue-specific risk coefficients and were used to derive patientspecific effective risk. The thirty patients had normalized effective dose of 3.7-10.4 mSv/100 mAs and normalized effective risk of 0.5-5.8 cases/1000 exposed persons/100 mAs. Normalized lung dose and risk of lung cancer correlated strongly with average chest diameter (correlation coefficient: r = -0.98 to -0.99). Normalized effective risk also correlated strongly with average chest diameter (r = -0.97 to -0.98). These strong correlations can be used to estimate patient-specific dose and risk prior to or after an imaging study to potentially guide healthcare providers in justifying CT examinations and to guide individualized protocol design and optimization.

Dynamic X-ray imagers require large surface, fast and highly sensitive X-ray absorbers and dedicated readout electronics. Monocrystalline photoconductors offer the sensitivity, speed, and MTF performances. Polycristalline photoconductors offer the large surface at a moderate cost. The challenge for them is to maintain the first performances at a compatible level with the medical applications requirements. This work has been focused on polycristalline CdTe grown by Close Space Sublimation (CSS) technique. This technique offers the possibility to grow large layers with a high material evaporation yield. This paper presents the results obtained with an image demonstrator using 350μm thick CdTe_css layers coupled to a CMOS readout circuit with Indium bumping. The present demonstrator has 200 x 200 pixels, with a pixel pitch of 75μm ×75μm. A total image surface of 15mm × 15mm has then been obtained. The ASIC works in an integration mode, i.e. each pixel accumulates the charges coming from the CdTe layer on a capacitor, converting them to a voltage. Single images as well as video sequences have been obtained. X-ray performance at 16 frames per second rate is measured. In particular a readout noise of 0.5 X ray, an MTF of 50% at 4 lp/mm and a DQE of 20% at 4lp/mm and 600 nGy are obtained. Although present demonstrator surface is moderate, it demonstrates that high performance can be expected from this assembly concept and its interest for medical applications.

We present an analysis of output referred pixel electronic noise as a function of position in the active matrix array for both active and passive pixel architectures. Three different noise sources for Active Pixel Sensor (APS) arrays are considered: readout period noise, reset period noise and leakage current noise of the reset TFT during readout. For the state-of-the-art Passive Pixel Sensor (PPS) array, the readout noise of the TFT switch is considered. Measured noise results are obtained by modeling the array connections with RC ladders on a small in-house fabricated prototype. The results indicate that the pixels in the rows located in the middle part of the array have less random electronic noise at the output of the off-panel charge amplifier compared to the ones in rows at the two edges of the array. These results can help optimize for clearer images as well as help define the region-of-interest with the best signal-to-noise ratio in an active matrix digital flat panel imaging array.

Contrast-enhanced digital subtraction mammography relies on the growth of new blood vessels (i.e. tumor angiogenesis) during the development of cancer. The growth accompanies an increase in tumor cell population to provide sufficient materials for cell proliferation. Since cancers will accumulate an injected contrast agent more than other tissues, it is possible to use one of several methods to enhance the area of lesions and remove the contrast of normal tissue. Large area flat panel detectors may be used for contrast-enhanced mammography wherein the subtraction of two acquired images is used to create the resulting enhanced image. Existing methods include temporal subtraction and dual energy subtraction, however these methods suffer from artifacts due to patient motion between the registration of images to be subtracted. In this paper we propose using a multilayer flat panel detector for contrast-enhanced digital subtraction mammography. The detector is designed to acquire both images simultaneously, thus avoiding motion artifacts in the resulting subtracted image. We study the multilayer detector design and examine the optimal weight factor and the signal difference to noise ratio. We find that the multilayer detector has the potential for energy discrimination, and thus the ability to be used for contrast-enhanced digital subtraction mammography.

A 2-TFT current-programmed, current-output active pixel sensor in amorphous silicon (a-Si:H) technology is introduced for digital X-ray imaging, and in particular, for mammography tomosynthesis and fluoroscopy. Pixel structure, operation and characteristics are presented. The proposed APS circuit was fabricated and assembled using an in-house bottom gate inverted staggered a-Si:H thin film transistor (TFT) process. Lifetime, transient performance as well as sensitivity to temperature measurements were carried out. An off-panel current amplifier with double sampling capability required for 1/f noise reduction is proposed and implemented in CMOS 0.18 micron technology. The results are promising and demonstrate that the proposed APS compensates for electrical and thermal stress causing shift in the threshold voltage of a-Si TFTs.

Resolution of positron emission tomography (PET) systems benefits from information about depth of interaction (DOI) within scintillation crystals, particularly in small bore scanners or parallel plate detectors. In this investigation, the ability of the dual-ended readout detector module configuration to resolve DOI and crystal index was evaluated for a variety of detector pitches and light guide thicknesses to validate the dual-ended readout method. Experimental results with oneto- one coupling between saw-cut 2mm pitch LYSO scintillation crystals and silicon photomultipliers (SiPMs) achieved 2.1 mm DOI resolution. Monte Carlo simulations were used to investigate the effect of larger detector pitches and varied light guide thickness on the crystal index identification accuracy and DOI resolution for a pixilated crystal array in dual-ended readout configuration. It is reported that the accuracy in identifying a 2 mm scintillation crystal was >80% for detector pitches < 6 mm and that DOI resolution was < 2 mm for all detector pitches and light guide thicknesses.

The new Solid State X-Ray Image Intensifier (SSXII) has the unique ability to operate in single photon counting (SPC) mode, with improved resolution, as well as in traditional energy integrating (EI) mode. The SSXII utilizes an electron-multiplying CCD (EMCCD), with an effective pixel size of 32μm, which enables variable signal amplification (up to a factor of 2000) prior to digital readout, providing very high-sensitivity capabilities. The presampled MTF was measured in both imaging modes using the standard angulated-slit method. A measured detector entrance exposure of 24μR per frame was used to provide approximately 0.8 interaction events per pixel in the 10μm-wide slit area. For demonstration purposes, a simple thresholding technique was used to localize events in SPC mode and a number of such frames were summed to provide an image with the same total exposure used for acquiring the EI image. The MTF for SPC mode, using a threshold level of 15% of the maximum 12-bit signal and 95% of the expected events, and for EI mode (in parentheses) was 0.67 (0.20), 0.37 (0.07), 0.20 (0.03), and 0.11 (0.01) at 2.5, 5, 7.5, and 10 cycles/mm, respectively. Increasing the threshold level resulted in a corresponding increase in the measured SPC MTF and a lower number of detected events, indicating a tradeoff between resolution and count efficiency is required. The SSXII in SPC mode was shown to provide substantial improvements in resolution relative to traditional EI mode, which should benefit applications that have demanding spatial resolution requirements, such as mammography.

Increasing the spatial resolution of current multislice clinical CT system is always desirable. However, further resolution improvement by reducing the pixel pitch or the aperture of the detector elements is difficult because of the tradeoff between the pixel size and dose level. In this paper, we demonstrate a methodology for improving spatial resolution of a clinical multislice CT without reducing the detector element size. The flying focal spot (i.e. electron beam wobbling) technique is used to increase the data sampling rate for in-plane (x-y) and z-axis scan acquisitions. In order to reduce the number of focal spot positions to achieve a certain spatial resolution, a super resolution technique using projections onto convex sets (POCS) is applied here to improve projection raw data sampling with reduced number of focal spot positions. The results indicate that it is possible to significantly increase spatial resolution on current multislice clinical CT systems without reducing the detector element size. In absence of noise, super resolution algorithms employing iterative regularization, such as POCS, can reduce the required number of focal spot positions. Thus, technical requirements on the multislice CT systems, such as rotation time and number of projections per rotation, can be much relaxed. However, noise reduction methods and methods of reducing projections per rotation, such as compressed sensing, are needed to work with super resolution technique to keep the radiation exposure from exceeding the current limit of clinical multislice CT.

This is the first report on a new fast statistical iterative reconstruction algorithm for conebeam with a circular source trajectory, accelerated by InstaRecon's fast O(N³logN) hierarchical cone beam backprojection1 and reprojection algorithms. We report on the results of image quality and run-time comparisons with iterative algorithms based on conventional backprojection and reprojection. We demonstrate that the iterative algorithm introduced here can provide image quality indistinguishable from an iterative algorithm using conventional BP/RP operators, while providing almost a 10x speedup in reconstruction rates. Combining the 10x algorithmic acceleration with additional hardware acceleration by FPGA, Cell, or GPU implementation, this work indicates the feasibility of iterative reconstruction algorithms for dose reduction and image quality improvement in routine CT practice, at competitive speeds and affordable cost.

Typical cupping correction methods are pre-processing methods which require either pre-calibration measurements or simulations of standard objects to approximate and correct for beam hardening and scatter. Some of them require the knowledge of spectra, detector characteristics, etc. The aim of this work was to develop a practical histogram-driven cupping correction (HDCC) method to post-process the reconstructed images. We use a polynomial representation of the raw-data generated by forward projection of the reconstructed images; forward and backprojection are performed on graphics processing units (GPU). The coefficients of the polynomial are optimized using a simplex minimization of the joint entropy of the CT image and its gradient. The algorithm was evaluated using simulations and measurements of homogeneous and inhomogeneous phantoms. For the measurements a C-arm flat-detector CT (FD-CT) system with a 30×40 cm² detector, a kilovoltage on board imager (radiation therapy simulator) and a micro-CT system were used. The algorithm reduced cupping artifacts both in simulations and measurements using a fourth-order polynomial and was in good agreement to the reference. The minimization algorithm required less than 70 iterations to adjust the coefficients only performing a linear combination of basis images, thus executing without time consuming operations. HDCC reduced cupping artifacts without the necessity of pre-calibration or other scan information enabling a retrospective improvement of CT image homogeneity. However, the method can work with other cupping correction algorithms or in a calibration manner, as well.

Computed tomography (CT) streak artifacts caused by metal implants have long been recognized as a problem that limits various applications of CT imaging. An effective and robust algorithm is highly desirable to minimize metal artifacts and achieve clinically acceptable CT images. In this work, the raw projection data is viewed as "incomplete" in the presence of metal shadows. Shape and location of metal objects are automatically identified and used as prior knowledge for accurate segmentation of metal shadows in projection domain. An iterative algorithm based on constrained optimization is then used for the image reconstruction. This algorithm minimizes a quadratic penalized smoothness measure function of the image, subject to the constraint that the estimated projection data is within a specified tolerance of the available metal-shadow-excluded projection data, with image non-negativity enforced. The constrained minimization problem is optimized through the combination of projection onto convex sets (POCS) and steepest gradient descent of the smoothness measure objective. Digital phantom study shows that the proposed constrained optimization algorithm has superior performance in reducing metal artifacts, suppressing noise and improving soft-tissue visibility. Some comparisons are performed with the filtered-back-projection (FBP), FDK, POCS and constrained optimization with total-variation (TV) objective. Although the algorithm is presented in the context of metal artifacts, it can be generated to image reconstruction from incomplete projections caused by limited angular range or low angular sampling rate in both 2D and 3D cases.

Differential phase contrast computed tomography (DPC-CT) is a novel X-ray imaging method that uses the wave properties of imaging photons as the contrast mechanism. It has been demonstrated that differential phase contrast images can be obtained using either synchrotron radiation or a conventional X-ray tube and a Talbot- Lau-type interferometer. These data acquisition systems offer only a limited field of view and thus, are prone to data truncation. In this work, we demonstrated that a small region of interest (ROI) of a large object can be accurately and stably reconstructed using fully truncated projection datasets provided that a priori information on electron density is known inside the ROI. The method reconstructs an image iteratively to satisfy a group of physical conditions using a projection onto convex set (POCS) algorithm. This POCS algorithm is validated using numerical simulations.

Coronary CT Angiography (CTA) is limited in patients with calcified plaque and stents. CTA is unable to confidently differentiate fibrous from lipid plaque. Fast switched dual energy CTA offers certain advantages. Dual energy CTA removes calcium thereby improving visualization of the lumen and potentially providing a more accurate measure of stenosis. Dual energy CTA directly measures calcium burden (calcium hydroxyapatite) thereby eliminating a separate non-contrast series for Agatston Scoring. Using material basis pairs, the differentiation of fibrous and lipid plaques is also possible. Patency of a previously stented coronary artery is difficult to visualize with CTA due to resolution constraints and localized beam hardening artifacts. Monochromatic 70 keV or Iodine images coupled with Virtual Non-stent images lessen beam hardening artifact and blooming. Virtual removal of stainless steel stents improves assessment of in-stent re-stenosis. A beating heart phantom with 'cholesterol' and 'fibrous' phantom coronary plaques were imaged with dual energy CTA. Statistical classification methods (SVM, kNN, and LDA) distinguished 'cholesterol' from 'fibrous' phantom plaque tissue. Applying this classification method to 16 human soft plaques, a lipid 'burden' may be useful for characterizing risk of coronary disease. We also found that dual energy CTA is more sensitive to iodine contrast than conventional CTA which could improve the differentiation of myocardial infarct and ischemia on delayed acquisitions. These phantom and patient acquisitions show advantages with using fast switched dual energy CTA for coronary imaging and potentially extends the use of CT for addressing problem areas of non-invasive evaluation of coronary artery disease.

Spectral Computed Tomography (Spectral CT), and in particular fast kVp switching dual-energy computed tomography, is an imaging modality that extends the capabilities of conventional computed tomography (CT). Spectral CT enables the estimation of the full linear attenuation curve of the imaged subject at each voxel in the CT volume, instead of a scalar image in Hounsfield units. Because the space of linear attenuation curves in the energy ranges of medical applications can be accurately described through a two-dimensional manifold, this decomposition procedure would be, in principle, limited to two materials. This paper describes an algorithm that overcomes this limitation, allowing for the estimation of N-tuples of material-decomposed images. The algorithm works by assuming that the mixing of substances and tissue types in the human body has the physicochemical properties of an ideal solution, which yields a model for the density of the imaged material mix. Under this model the mass attenuation curve of each voxel in the image can be estimated, immediately resulting in a material-decomposed image triplet. Decomposition into an arbitrary number of pre-selected materials can be achieved by automatically selecting adequate triplets from an application-specific material library. The decomposition is expressed in terms of the volume fractions of each constituent material in the mix; this provides for a straightforward, physically meaningful interpretation of the data. One important application of this technique is in the digital removal of contrast agent from a dual-energy exam, producing a virtual nonenhanced image, as well as in the quantification of the concentration of contrast observed in a targeted region, thus providing an accurate measure of tissue perfusion.

Dual energy CT (DECT) provides material-selective CT images by acquiring the object of interest with two different x-ray spectra, a low and a high energy spectrum. Today, two techniques to process the rawdata are in use: Image-based DECT reconstructs the low and the high energy data separately and then performs a linear combination of the images to yield the desired material-selective images. This method can only provide a first order approximation of the true material decomposition and it will not be able to remove higher order beam hardening artifacts from the images. By contrast, rawdata-based DECT naturally deals with higher order effects and is therefore the better way to go. However, rawdata-based DECT requires the same line integrals to be available for both scans (consistent scans). This requirement may not be met for CT scanners that are available today. To handle the material decomposition of DECT from inconsistent scans (i.e. non-overlapping rays for each measured spectrum) a material decomposition algorithm (MDIR) that allows for different scan trajectories and scan geometries for the low and the high energy scan was developed and evaluated. The results of our iterative algorithm are comparable to those obtained by a rawdata-based approach. However, conventional rawdata-based approaches are often not applicable since inconsistent rays are acquired. It should be noted that MDIR can be extended to scans with more than two different spectra and to decompositions into more than two basis functions in a straightforward way.

Dual-energy CT has attracted much attention in recent years. Most recently, a fast-kVp switching (FKS) dual-energy method has been presented with clinical and phantom results to demonstrate its efficacy. The purpose of our study was to quantitatively compare the CTDIW of FKS and routine CT exams under the body and head conditions. For a fair comparison, the low contrast detectability (LCD) was matched before measuring dose. In FKS protocols, an x-ray generator switch rapidly between 140kVp and 80kVp in adjacent views, and the effective tube current is around 600mA. In addition to the tube voltage and current, the flux ratio between high and low kVp is optimized by asymmetric sampling of 35%-65%. The head and body protocols further differ by the gantry speed (0.9sec/1.0sec) and type of bowtie filter (head/body). For baseline single-energy, we followed the IEC standard head and body protocols (120kV, 1sec, 5mm) but iteratively adjusted the tube current (mA) in order to match the LCD. CTDIW was measured using either a 16 cm (for head scanning) or a 32 cm (for body scanning) PMMA phantom of at least 14 cm in length. The LCD was measured using the water section of Catphan 600. To make the study repeatable, the automated statistical LCD measurement tool available on GE Discovery CT750 scanner was used in this work. The mean CTDIW for the head and body single-energy acquisitions were 57.5mGy and 29.2mGy, respectively. The LCD was measured at 0.45% and 0.42% (slice thickness=5mm, object size=3mm, central 4 images), respectively. The average CTDIW for FKS head and body scans was 70.4mGy and 33.4mGy, respectively, at the optimal monochromatic energy of 65 keV. The corresponding LCD was measured at 0.45% and 0.43%, respectively. This demonstrates that, with matching LCD, CTDIW of FKS is comparable to that of routine CT exams under head and body conditions.

Recent publications emphasize the benefits of quantum-counting applied to the field of Computed Tomography (CT). We present a research prototype scanner with a CdTe-based quantum-counting detector and 20 cm field-of-view (FOV). As of today there is no direct converter material on the market able to operate reliably in the harsh high-flux regime of clinical CT scanners. Nevertheless, we investigate the CT imaging performance that could be expected with high-flux capable material. Therefore we chose pixel sizes of 0.05 mm², a good compromise between high-flux counting ability and energy resolution. Every pixel is equipped with two energy threshold counters, enabling contrast-optimization and dual-energy scans. We present a first quantitative analysis of contrast measurements, in which we limit ourselves to a low-flux scenario. Using an Iodine-based contrast agent, we find 17% contrast enhancement at 120 kVp, compared to energy-integrating CT. In addition, the general dual-energy capability was confirmed in first measurements. We conclude our work by demonstrating good agreement of measurement results and detailed CT-system simulations.

In Computed Tomography applications a major opportunity has been identified in the exploitation of the spectral information inherently available due to the polychromatic emission of the X-ray tube. Current CT technology based on indirect-conversion and integrating-mode detection can be used to some extent to distinguish the two predominant physical causes of energy-dependent attenuation (photo-electric effect and Compton effect) by using dual-energy techniques, e.g. kVp switching, dual-source or detector stacking. Further improvements can be achieved by transitioning to direct-conversion technologies and counting-mode detection, which inherently exhibits a better signal-to-noise ratio. Further including energy discrimination, enables new applications, which are not feasible with dual-energy techniques, e.g. the possibility to discriminate K-edge features (contrast agents, e.g. Gadolinium) from the other contributions to the x-ray attenuation of a human body. The capability of providing energy-resolved information with more than two different measurements is referred to as Spectral CT. To study the feasibility of Spectral CT, an energy-resolving proprietary photon counting ASIC (ChromAIX) has been designed to provide high count-rate capabilities while offering energy discrimination. The ChromAIX ASIC consists of an arrangement of 4 by 16 pixels with an isotropic pitch of 300 μm. Each pixel contains a number of independent energy discriminators with their corresponding 12-bit counters with continuous read-out capability. Observed Poissonian count-rates exceeding 10 Mcps (corresponding to approximately 27 Mcps incident mean Poisson rate) have been experimentally validated through electrical characterization. The measured noise of 2.6 mV_RMS (4 keV FWHM) adheres to specifications. The ChromAIX ASIC has been specifically designed to support direct-converting materials CdZnTe and CdTe.

Dual energy CT using a fast kVp switching technique and the standard filtered back projection (FBP) image reconstruction method has recently been studied. With conventional FBP methods, high slew rates are required for acceptable image reconstruction with high image quality. However, high slew rates also require hardware changes to enable data acquisition. In this work, we aim at studying the necessary slew rate for dual energy CT imaging provided that the PICCS algorithm is used for image reconstruction. The results demonstrate that a slew rate of 7.5 kV / view (assuming 2,000 views were collected over 360^o with a 60 kVp energy separation) was sufficient for dual energy imaging using PICCS.

A technique for temporal resolution improvement using prior image constrained compressed sensing (TRI-PICCS) in multi-detector computed tomography (MDCT) cardiac imaging is proposed. In this work, the performance of TRIPICCS was studied using a hybrid phantom which consists of realistic cardiac anatomy and objects moving with designed trajectories. Several simulated moving vessels were added to different locations in the heart. Different motion directions and simulated heart rates were investigated using half of the projection data of the short-scan angular range in TRI-PICCS. Different angular ranges of projection data were also investigated in TRI-PICCS to evaluate the highest achievable temporal resolution. The results showed that the temporal improvement of TRI-PICCS is independent of the locations of the moving objects and motion directions. The motion artifacts at 100 bmp simulated heart rate can be significantly improved using TRI-PICCS compared with conventional filtered back projection (FBP). The minimum angular range requirement of TRI-PICCS is about 90°, corresponding to a temporal resolution improvement factor of 2.6 compared with the standard short-scan FBP reconstruction.

In medical imaging, image background is often defined by zero signal. Moreover, in medical images the background area - or conversely, the spatial support (the extent of the non-zero part of the image) - is often known a priori or can be easily estimated. For example, support information can be estimated from the low-resolution "scout" images typically acquired during pre-scan localization in both MRI and CT. In dynamic scans, object support in a single time-frame is often obtainable from a prior time frame, or from a composite image formed from data from multiple time frames. In this work, incorporation of either complete or partial a priori knowledge of object spatial support into the compressive sensing (CS) framework is investigated. Following development of the augmented reconstruction model, examples of support-constrained CS reconstruction of phantom and MR images under both exact and inexact support definitions are given. For each experiment, the straightforward incorporation of the proposed spatial support constraint into the standard CS model was shown to both significantly accelerate reconstruction convergence and yield a lower terminal RMSE compared to a conventional CS reconstruction. The proposed augmented reconstruction model was also shown to be robust to inaccuracies in the estimated object support.

Conventional dynamic PET studies estimate pharmacokinetic parameters using a two-step procedure of first reconstructing the spatial activity volume for each temporal frame independently before applying a pharmacokinetic model to the resulting spatio-temporal activity distribution. This indirect procedure leads to low SNR due to using only a subset of the temporal data when reconstructing each image. Our work concentrates on the estimation of parameters directly from the (raw or pre-corrected) dPET temporal projections. We present here a one-step direct pharmacokinetic algorithm based on the Ordered Subset (OS) Weighted Least Squares (WLS) iterative estimation algorithm. We explicitly incorporate a priori temporal information by modelling the Time Activity Curves (TACs) as a sum of exponentials convolved with an Input Function. Our OS-WLS-PK algorithm is appropriate for both 3D projection data which has been Fourier Rebinned into 2D slices, as well as when the data has been pre-corrected for attenuation, randoms and scatter. The main benefit of spectral analysis applied to dynamic PET reconstruction is that no particular pharmacokinetic model needs to be specified a priori, with only the input function needing to be sampled at scan time. We test our algorithm on highly realistic SORTEO generated data and show it leads to more accurate parameter estimates than when conventional graphical methods are used.

We evaluate the energy dependent noise and bias properties of monoenergetic images synthesized from dual-energy CT (DECT) acquisitions used to estimate attenuation coefficients at PET or SPECT energies. This is becoming more relevant with the increased used of quantitative imaging by PET/CT and SPECT/CT. There are, however, variations in the noise and bias properties of synthesized monoenergetic images as a function of energy. We used analytic approximations and simulations to estimate the bias and noise of synthesized monoenergetic images of a water-filled cylinder from 10 to 525 keV. The dual-energy spectra were based on the GE Lightspeed VCT scanner at 80 and 140 kVp. Both analytic calculations and simulations for increasing energy the relative noise plateaued near 140 keV (i.e. SPECT with ^99mTc), and then remained constant with increasing energy up to 511 keV and beyond (i.e. PET). If DECT is being used for attenuation correction at higher energies, there is a noise amplification that is dependent on the energy of the synthesized monoenergetic image of linear attenuation coefficients. For SPECT and PET imaging the bias and noise levels of DECT based attenuation correction is unlikely to affect image quality.

The high flexibility of C-arm flat-detector computed tomography (FDCT) is used in a volume of interest (VOI) imaging method to handle the challenges of increasing spatial resolution, reducing noise and saving dose. A low-dose overview scan of the object and a high-dose scan of an arbitrary VOI are combined. The first scan is adequate for orientation to select the VOI and the second scan assures high image quality in the VOI. The combination is based on a forward projection of the reconstructed overview volume and the measured VOI data in the raw data domain. Differences in the projection values are matched before a standard Feldkamp-type reconstruction is performed. In simulations, spatial resolution, noise and contrast detectability were evaluated. Measurements of an anthropomorphic phantom were used to validate the proposed method for realistic application. In Monte Carlo dose simulations the dose reduction potential was investigated. By combination of the two scans an image is generated which covers the whole object and provides the actual VOI at high image quality. Spatial resolution was increased whereas noise was decreased from outside to inside the VOI, e.g. for the simulations from 0.8 lp/mm to 3.0 lp/mm and from 39 HU to 18 HU, respectively. Simultaneously, the cumulative dose for this two-scan procedure was significantly reduced in comparison to a conventional high dose scan, e.g. for the performed simulations and measurements by about 95 %. The proposed VOI approach offers significant benefits with respect to high-resolution and low-contrast imaging of a VOI at reduced dose.

We have described methods that allow highly accelerated MRI using under-sampled acquisitions and constrained reconstruction. One is a hybrid acquisition involving the constrained reconstruction of time dependent information obtained from a separate scan of longer duration. We have developed reconstruction algorithms for DSA that allow use of a single injection to provide the temporal data required for flow visualization and the steady state data required for construction of a 3D-DSA vascular volume. The result is time resolved 3D volumes with typical resolution of 512³ at frame rates of 20-30 fps. Full manipulation of these images is possible during each stage of vascular filling thereby allowing for simplified interpretation of vascular dynamics. For intravenous angiography this time resolved 3D capability overcomes the vessel overlap problem that greatly limited the use of conventional intravenous 2D-DSA. Following further hardware development, it will be also be possible to rotate fluoroscopic volumes for use as roadmaps that can be viewed at arbitrary angles without a need for gantry rotation. The most precise implementation of this capability requires availability of biplane fluoroscopy data. Since the reconstruction of 3D volumes presently suppresses the contrast in the soft tissue, the possibility of using these techniques to derive complete indications of perfusion deficits based on cerebral blood volume (CBV), mean transit time (MTT) and time to peak (TTP) parameters requires further investigation. Using MATLAB post-processing, successful studies in animals and humans done in conjunction with both intravenous and intra-arterial injections have been completed. Real time implementation is in progress.

C-arm CT is used in neurovascular interventions where a large flat panel detector is used to acquire cone-beam projection data. In this case, data truncation problems due to the limited detector size are mild. When the cone beam CT method is applied to cardiac interventions severe data truncation artifacts reduce the clinical utility of the reconstructions. However, accurate reconstruction is still possible given a priori knowledge of the reconstruction values within a small region inside the FOV. Several groups have studied the case of the interior problem where data is truncated from all views. In this paper, we applied these new mathematical discoveries to C-arm cardiac cone-beam CT to demonstrate that accurate image reconstruction may be achieved for cardiac interventions. The method is applied to iteratively reconstruct the image volume such that it satisfies several physical conditions. In this work, the algorithm is applied to data from in-vivo cardiac canine studies collected using a clinical C-arm system. It is demonstrated that the algorithm converges well to the reconstruction values of non-truncated data reconstructed using the FDK algorithm. Furthermore, proper convergence is achieved by using only an estimate of the average value within a subregion as a priori information (i.e. the exact value at each pixel in the a priori region need not be known). Two methods for obtaining a priori information are compared.

A dual modality SPECT-CT prototype dedicated to uncompressed breast imaging (mammotomography) has been developed. The CT subsystem incorporates an ultra-thick K-edge filtration technique producing a quasi-monochromatic x-ray cone beam to optimize the dose efficiency for uncompressed breast tomography. We characterize the absorbed dose to the breast under normal tomographic cone beam image acquisition protocols using both TLD measurements and ionization chamber-calibrated radiochromic film. Geometric and anthropomorphic breast phantoms are filled with 1000mL of water and oil to simulate different breast compositions and varying object shapes having density bounds of 100% glandular and fatty breast compositions, respectively. Doses to the water filled geometric and anthropomorphic breast phantoms for a tomographic scan range from 1.3-7.3mGy and 1.7-6.3mGy, respectively, with a mean whole-breast dose of 4.5mGy for the water-filled anthropomorphic phantom. Measured dose distribution trends indicate lower doses in the center of the breast phantoms towards the chest wall along with higher doses near the peripheries and nipple regions. Measured doses to the oil-filled phantoms are consistently lower across all volume shapes (mean dose, 3.8mGy for the anthropomorphic breast). Results agree with Monte Carlo dose estimates generated for uncompressed breast imaging and illustrate the advantages of using the novel K-edge filtered beam to minimize absorbed dose to the breast during fully-3D imaging.

The impact of the system parameters of the modulator on X-ray scatter correction using primary modulation is studied and an optimization of the modulator design is presented. Recently, a promising scatter correction method for X-ray computed tomography (CT) that uses a checkerboard pattern of attenuating blockers (primary modulator) placed between the X-ray source and the object has been developed and experimentally verified. The blocker size, d, and the blocker transmission factor, α, are critical to the performance of the primary modulation method. In this work, an error caused by aliasing of primary whose magnitude depends on the choices of d and α, and the scanned object, is set as the object function to be minimized, with constraints including the X-ray focal spot, the physical size of the detector element, and the noise level. The optimization is carried out in two steps. In the first step, d is chosen as small as possible but should meet a lower-bound condition. In the second step, α should be selected to balance the error level in the scatter estimation and the noise level in the reconstructed image. The lower bound of d on our tabletop CT system is 0.83 mm. Numerical simulations suggest 0.6 < α < 0.8 is appropriate. Using a Catphan 600 phantom, a copper modulator (d = 0.89 mm, α = 0.70) expectedly outperforms an aluminum modulator (d = 2.83 mm, α = 0.90). With the aluminum modulator, our method reduces the average error of CT number in selected contrast rods from 371.4 to 25.4 Hounsfield units (HU) and enhances the contrast to noise ratio (CNR) from 10.9 to 17.2; when the copper modulator is used, the error is further reduced to 21.9 HU and the CNR is further increased to 19.2.

We have previously reported the development of a dynamic micro-CT scanner with a stationary mouse bed using a compact carbon nanotube (CNT) field emission x-ray tube and preliminary results on its utility for prospectively gated cardiac imaging. In this paper we report the recent progress in improving the performance characteristics of this scanner. Through optimization of the CNT cathode, the stable emission current has been increased. The output power of the CNT x-ray source has reached ~100W peak power at 100μm focal spot size. The higher flux enables improvement of the xray energy spectrum to minimize the beam hardening effect and increasing the system temporal resolution by using shorter x-ray exposure time. The scanner's temporal resolution has been increased to ~10 msec, which is sufficient for high-resolution micro-CT imaging of mouse heart and lung under free-breathing setting. The spatial resolution is maintained at 6.2 lp per mm at 10% system MTF. The nanotube micro-CT scanner's application in mouse cardiac imaging has been demonstrated with high-resolution (80 μm and 15 msec) micro-CT of the mouse heart under freebreathing setting.

The purpose of this paper is to present a new image reconstruction algorithm for dynamic data, termed non-convex prior image constrained compressed sensing (NC-PICCS). It generalizes the prior image constrained compressed sensing (PICCS) algorithm with the use of non-convex priors. Here, we concentrate on perfusion studies using computed tomography examples in simulated phantoms (with and without added noise) and in vivo data, to show how the NC-PICCS method holds potential for dramatic reductions in radiation dose for time-resolved CT imaging. We show that NC-PICCS can provide additional undersampling compared to conventional convex compressed sensing and PICCS, as well as, faster convergence under a quasi-Newton numerical solver.

Adaptive Statistical Iterative Reconstruction (ASIR) is a new imaging reconstruction technique recently introduced by General Electric (GE). This technique, when combined with a conventional filtered back-projection (FBP) approach, is able to improve the image noise reduction. To quantify the benefits provided on the image quality and the dose reduction by the ASIR method with respect to the pure FBP one, the standard deviation (SD), the modulation transfer function (MTF), the noise power spectrum (NPS), the image uniformity and the noise homogeneity were examined. Measurements were performed on a control quality phantom when varying the CT dose index (CTDI_vol) and the reconstruction kernels. A 64-MDCT was employed and raw data were reconstructed with different percentages of ASIR on a CT console dedicated for ASIR reconstruction. Three radiologists also assessed a cardiac pediatric exam reconstructed with different ASIR percentages using the visual grading analysis (VGA) method. For the standard, soft and bone reconstruction kernels, the SD is reduced when the ASIR percentage increases up to 100% with a higher benefit for low CTDIvol. MTF medium frequencies were slightly enhanced and modifications of the NPS shape curve were observed. However for the pediatric cardiac CT exam, VGA scores indicate an upper limit of the ASIR benefit. 40% of ASIR was observed as the best trade-off between noise reduction and clinical realism of organ images. Using phantom results, 40% of ASIR corresponded to an estimated dose reduction of 30% under pediatric cardiac protocol conditions. In spite of this discrepancy between phantom and clinical results, the ASIR method is as an important option when considering the reduction of radiation dose, especially for pediatric patients.

Mechanical stability of a medical LINear ACcelerator (LINAC), particularly the quality of the gantry, collimator and table rotations and the accuracy of the isocenter position, are crucial for the radiation therapy process, especially in stereotactic radio surgery and in Image Guided Radiation Therapy (IGRT) where this mechanical stability is perturbed due to the additional weight the kV x-ray tube and detector. In this paper, we present a new method to evaluate a software which is used to perform an automatic measurement of the "size" (flex map) and the location of the kV and the MV isocenters of the linear accelerator. The method consists of developing a complete numerical 3D simulation of a LINAC and physical phantoms in order to produce Electronic Portal Imaging Device (EPID) images including calibrated distortions of the mechanical movement of the gantry and isocenter misalignments.

An angular parameterization of parallel Radon projections referred to in this paper as ψ-parameterization is discussed in relevance to the efficiency of reconstruction from fan data. The fact that the ψ-parameterization coincides with the equiangular fan beam parameterization allows us to develop a simple and efficient approach useful for the reconstruction from fan data. Within this approach parallel projections are approximated by groups of semi-parallel rays. The reconstruction is carried out directly, i.e. without any modification of original data, at the speed which is comparable or even higher than that of the parallel Filtered Back Projection (FBP) algorithm.

Reconstruction algorithms based on PI-line or Chord are active subject in CBCT. Among them back-projection filtered (BPF) reconstruction algorithm has obvious influence for its exact reconstruction results and less computations especially in selected volume of interesting (VOI) regions. However, the selecting and sampling method of PI-line segment can directly affect the quality of reconstructed images. In this paper, we proposed a general PI-line selecting scheme to reconstruct VOI regions by using BPF algorithm, which mainly based on the relationship between reconstructed coordinate and PI-line coordinate. The proposed scheme is applicable for GPU accelerated back-projection filtered reconstruction.

The limited angle problem is a well-known problem in computed tomography. It is caused by missing data over a certain angle interval, which make an inverse Radon transform impossible. In daily routine this problem can arise for example in tomosynthesis, C-arm CT or dental CT. In the last years there has been a big development in the field of compressed sensing algorithms in computed tomography, which deal very good with incomplete data. The most popular way is to integrate a minimal total variation norm in form of a cost function into the iteration process. To find an exact solution of such a constrained minimization problem, computationally very demanding higher order algorithms should be used. Due to the non perfect sparsity of the total variation representation, reconstructions often show the so called staircase effect. The method proposed here uses the solutions of the iteration process as an estimation for the missing angle data. Compared to a pure compressed sensing-based algorithm we reached much better results within the same number of iterations and could eliminate the staircase effect. The algorithm is evaluated using measured clinical datasets.

We developed an embossed radiography system utilizing single- and dual-energy subtractions for decreasing the absorption contrast of unnecessary regions, and contrast resolution of a target region was increased using image-shifting subtraction and a linear-contrast system in a flat panel detector (FPD). To carry out embossed radiography, we developed a computer program for two-dimensional subtraction, and a conventional x-ray generator with a 0.5-mm-focus tube was used. Energy subtraction was performed at tube voltages of 42.5 and 70.0 kV, a tube current of 1.0 mA, and an x-ray exposure time of 5.0 s. Embossed radiography was achieved with cohesion imaging by use of the FPD with pixel sizes of 48 ×48 μm, and the shifting dimension of an object in the horizontal and vertical directions ranged from 48 to 144 μm. We obtained high-contrast embossed images of fine bones and coronary arteries approximately 100 μm in diameter.

The purpose of this study was to compare lesion detection in images reconstructed using standard filtered back projection (FBP) with those reconstructed using a new CT reconstruction algorithm called Iterative Reconstruction in Image Space (IRIS). Detection performance was experimentally measured using a 2- AFC software package that computes the lesion intensity corresponding to a detection accuracy of 92% (i.e., I92%). Abdominal images were acquired on a Siemens Somaton Definition Flash CT scanner and reconstructed at four slice thickness values ranging from 1.5 mm to 10 mm. Detection of three lesion sizes was investigated, whose diameters ranged from 5 mm to 10 mm. AFC experiments were performed using FBP and IRIS reconstructed images that were presented to observers in a random manner. For any lesion in a given image, we obtained an Enhancement Factor (EF) defined as the I92% using FBP divided by the corresponding I92% using IRIS. In 9 out of 12 paired results, EF values were significantly greater than 1.0, and in the remaining three cases, EF values were approximately 1.0. EF was independent of CT image slice thickness, with an average value of 1.17 ± 0.12. Values of EF increased with decreasing lesion size, and were about 20% greater for 5 mm lesions than 10 mm lesions. The results of this pilot study show that IRIS improved lesion detection compared to conventional FBP, with an average increase in signal to noise ratio of 17%. For the smallest lesions, improvements in signal to noise ratio approached 30%. Our results suggest that radiation dose reductions of one third might be achievable for abdominal imaging without any loss in signal to noise ratio.

3D computed tomography has been extensively studied and widely used in modern society. Although most manufacturers choose the filtered backprojection algorithm (FBP) for its accuracy and efficiency, iterative reconstruction methods have a significant potential to provide superior performance for incomplete, noisy projection data. However, iterative methods have a high computational cost, which hinders their practical use. Furthermore, regularization is usually required to reduce the effects of noise. In this paper, we analyze the use of the Simultaneous Algebraic Reconstruction Technique (SART) with total variation (TV) regularization. Additionally, graphics hardware is utilized to increase the speed of SART. NVIDIA's GPU and Compute Unified Device Architecture (CUDA) comprise the core of our computational platform. GPU implementation details, including ray-based forward projection and voxel-based backprojection are illustrated. Experimental results for high-resolution synthetic and real data are provided to demonstrate the accuracy and efficiency of the proposed framework.

Quantification of SPECT(Single Photon Emission Computed Tomography) images can be more accurate if correct segmentation of region of interest (ROI) is achieved. Segmenting ROI from SPECT images is challenging due to poor image resolution. SPECT is utilized to study the kidney function, though the challenge involved is to accurately locate the kidneys and bladder for analysis. This paper presents an automated method for generating seed point location of both kidneys using anatomical location of kidneys and bladder. The motivation for this work is based on the premise that the anatomical location of the bladder relative to the kidneys will not differ much. A model is generated based on manual segmentation of the bladder and both the kidneys on 10 patient datasets (including sum and max images). Centroid is estimated for manually segmented bladder and kidneys. Relatively easier bladder segmentation is followed by feeding bladder centroid coordinates into the model to generate seed point for kidneys. Percentage error observed in centroid coordinates of organs from ground truth to estimated values from our approach are acceptable. Percentage error of approximately 1%, 6% and 2% is observed in X coordinates and approximately 2%, 5% and 8% is observed in Y coordinates of bladder, left kidney and right kidney respectively. Using a regression model and the location of the bladder, the ROI generation for kidneys is facilitated. The model based seed point estimation will enhance the robustness of kidney ROI estimation for noisy cases.

A novel metal artifact reduction strategy including projection reformation, metal region segmentation, and metal shadow filling was proposed. Both metal region segmentation and shadow filling are critical steps to assure good artifact suppression results. This preliminary study evaluated the performance of two segmentation methods and three region filling methods on metal artifact reduction of clinical cases. Gradient-based threshold method (GBT) and dual-front active contour model-based method (DFAC) were utilized to segment metal implants from reformatted projections, Delaunay triangulation-based (DTB), anisotropic diffusion-based, and exemplar-based, interpolation methods were utilized to fill the metal shadows, respectively. The image quality was evaluated by a radiologist in terms of visual conspicuity of the bladder base, prostate, and rectum. Overall, the image quality and the conspicuity in some critical organs were significantly improved for all corrections. Compared to the GBT method, the DFAC method had more accurate segmentation, which resulted in better artifact suppression. The interpolation process does not guarantee the data consistency among projection views, which can introduce additional artifacts, especially for large metal objects. Although the DTB method produced the smoothest metal shadow interpolation results, which is considered the worst scenario according to the criterion of image restoration, it induced the least additional artifacts to the reconstructed images compared to the other two structure-saving methods. As such, region interpolation methods should follow the criterion to generate metal shadow data consistent with the CT acquisition geometry, which might be quite different from the general standard of image restoration in computer vision and image processing.

We have recently developed a locally-adaptive method for noise control in CT based upon bilateral filtering. Different from the previous adaptive filters, which were locally adaptive by adjusting the filter strength according to local photon statistics, our use of bilateral filtering in projection data incorporates a practical CT noise model and takes into account the local structural characteristics, and thus can preserve edge information in the projection data and maintain the spatial resolution. Despite the incorporation of the CT noise model and local structural characteristics in the bilateral filtering, the noise-resolution properties of the filtered image are still highly dependent on predefined parameters that control the weighting factors in the bilateral filtering. An inappropriate selection of these parameters may result in a loss of spatial resolution or an insufficient reduction of noise. In this work, we employed an adaptive strategy to modulate the bilateral filtering strength according to the noise-equivalent photon numbers determined from each projection measurement. We applied the proposed technique to head/neck angiographic CT exams, which had highly non-uniform attenuation levels during the scan. The results demonstrated that the technique can effectively reduce the noise and streaking artifacts caused by high attenuation, while maintaining the reconstruction accuracy in less attenuating regions.

Micro-CT enables convenient visualization and quantitative analysis of small animals and biological tissue samples. However, high-quality volume images in general require acquisition of cone-beam projection data from hundres of view angles. This prolonged imaging process limits system throughput and may cause potential radiation damage to the imaged objects. It is therefore desirable to have a technique which can generate volume images with satisfactory quality, but from a smaller amount of projection data. On the other hand, many objects subject to the micro-CT scans have sparse spatial distribution, and this sparcity could be exploited and incorporated as prior knowledge in innovative design of algorithms that are capable of reconstructuring images from few-view projection data. In this work we applied a new iterative algorithm based upon constrained total-variation minimization to reconstructing images from as few as five projections. Preliminary results suggest that the algorithm can yield potentially useful images from substantially less projection data than required by existing algorithms. This has practical implications of reducing scanning time and minimizing radiation damage to the imaged objects.

The purpose of this study was to quantify how reducing x-ray beam intensity (i.e., mAs) affects lesion detection performance in abdominal CT examinations. A simulation package (Syngo Explorer) was used to reconstruct 4-mm thick CT images of a patient undergoing a standard abdominal exam. Simulations were performed at four x-ray beam intensities of 100%, 70%, 50%, and 25%. Four observers were used to perform a series of two Alternate Forced Choice (2-AFC) experiments that measure the lesion contrast (I_92%) corresponding to a detection accuracy of 92%. Four lesion sizes were used ranging from 5 mm to 12 mm. Results were plotted as log(I_92%) versus log(mAs) to quantify how changes in x-ray intensity affect lesion detection, as well as log(I_92%) versus log(size) to generate contrast-detail curves. The fitted slope of noise in reconstructed images as a function of relative CT x-ray beam intensity was -0.25, which is about half the value of -0.5 expected for an ideal quantum noise limited imaging system. For lesion sizes between 5 mm and 10 mm, slopes of log(I92%) versus log(mAs) curves were very similar for all four observers, and ranged between -0.10 and -0.17. For 5 mm sized lesions, doubling the x-ray beam intensity improved detection performance by about 13%, whereas for 7 and 10 mm lesions, doubling the x-ray intensity improved detection performance by about 7%. For the 12 mm lesion there were no consistent patterns for all four readers, which may be related to the lack of a standardized viewing distance. The average slope for the four contrast detail curves was -0.41 ± 0.09, which is substantially less than the value of -1.0 predicted for an ideal observer operating with a quantum noise limited images. For our abdominal CT images, doubling of the lesion size resulted in improvements in lesion detection of ~ 30%.

In recent years, there has been a rapid research progress of molecular imaging technology. Many investigations in molecular imaging such as the nanoparticle applications in targeted drug delivery have been widely studied in several key small animal models. Various nanoparticles used as either the drug delivery carriers, imaging contrast mediums or target-specific therapeutic agents have established a novel research platform for biomedical related scientists and clinicians. Among these nanoparticles, gold nanoparticles have the unique non-toxic and stability properties. In this work, a commercially-available micro CT imaging system was used to specifically study the imaging properties for 15 nm spherical-shaped gold particles. Imaging properties were quantified by the CT numbers obtained from a series of photon energy levels in the micro CT scanner. We also compared the imaging results between gold nanoparticles and iodinated contrast medium to study the potential impact of gold nanoparticles served as the contrast agent.

Energy-integrating detection of x-ray sources is widely used by modern computed tomography (CT) scanners and has been an interesting research topic for the purpose of more accurately processing the data toward low-dose applications. While the energy-integrating detection can be described by a compound Poisson distribution, this work provides an alternative means to explicitly consider the Poisson statistics of the quanta and the energy spectrum of the x-ray generation. An exact solution for the first two orders of the compound Poisson statistics is presented. Given the energy spectrum of an x-ray source, the mean and variance of the measurement at any count-density level can be computed strictly. This solution can provide a quantitative measure on the condition under which an assumption of employing the most commonly-used independent identical distribution (i.i.d.), such as Gamma, Gaussian, etc, would be valid. A comparison study was performed to estimate the introduced errors of variance by using these substitute statistical functions to approximate the actual photon spectrum. The presented approach would further be incorporated in an adaptive noise treatment method for low-dose CT applications.

Today Cardiac imaging is one of the key driving forces for the research and development activities of Computed Tomography (CT) imaging. It requires high spatial and temporal resolution and is often associated with high radiation dose. The newly introduced ASIR technique presents an efficient method that offers the dose reduction benefits while maintaining image quality and providing fast reconstruction speed. This paper discusses the study of image quality of the ASIR technique for Cardiac CT imaging. Phantoms as well as clinical data have been evaluated to demonstrate the effectiveness of ASIR technique for Cardiac CT applications.

In this paper, a novel regularization approach for (non-statistical) iterative reconstruction is developed. In our implementation, the update equation of iterative reconstruction is based on Filtered Backprojection (FBP) and the solution is stabilized using nonlinear regularization priors. It is well known that the usage of nonlinear regularization priors can reduce image noise at the same time preserving image sharpness [1]. The final noise level can be adjusted by dedicated choice of regularization priors, regularization strength and the total number of iterations. In contrast to conventional CT using convolution kernels, image characteristics can not be further manipulated. This might cause artificial image texture. We present a new class of (non-local) 3D-regularization priors, which gives us control over image characteristics similar to that obtained with conventional CT convolution kernels. In addition, efficient noise reduction at constant sharpness is obtained. Due to the manipulation of the low-frequency components of the regularization filter, the filter is non-local. The regularization strength becomes a 3D-matrix with contrast-dependent entries, which gives us control over contrastdependent sharpness. The contrast edges are estimated using a 3D Laplacian kernel. High contrast edges get a low regularization weight and vice versa. We demonstrate the potential of noise reduction on basis of clinical CT data. Also, it is shown, that radiation exposure to the patient can be reduced by 60% in general purpose radiological CT applications and cardiac CT at the same time maintaining image quality. Moreover, for a 128-slice detector with 0.6 mm collimation, it is shown, that cone-beam and spiral artifacts caused by non-exact image reconstruction can be fairly removed. Putting all together our iterative reconstruction approach substantially improves image quality in cone-beam CT, and thus has the potential to enter routine clinical CT.

Current lung nodule size assessment methods typically rely on one-dimensional estimation of lesions. While new 3D volume assessment techniques using MSCT scan data have enabled improved estimation of lesion size, the effect of acquisition and reconstruction parameters on accuracy and precision of such estimation has not been adequately investigated. To characterize such dependencies, we scanned an anthropomorphic thoracic phantom containing synthetic nodules with different protocols, including various acquisition and reconstruction parameters. We also scanned the phantom repeatedly with the same protocol to investigate repeatability. The nodule's volume was estimated by a clinical lung analysis software package, LungVCAR. Accuracy (bias) and precision (variance) of the volume assessment were calculated across the nodules and compared between protocols via Generalized Estimating Equation analysis. Results suggest a strong dependence of accuracy and precision on dose level but little dependence on reconstruction thickness, thus providing possible guidelines for protocol optimization for quantitative tasks.

Imaging the basic functional unit (BFU) of small/medium animal - an organ's smallest assembly of diverse cells that functions like the organ itself - is of significance in pre-clinical research. A BFU is usually a spheroid with its horizontal, transverse and vertical radii equal to ~50 μm and its dimension is virtually the same in small/medium animal and human. Apparently, state-of-the-art diagnostic CT scanners can't image the BFU of small/medium animal directly because of its insufficient spatial resolution. Micro-CT is of sufficient spatial resolution, but its contrast, temporal and spectral resolutions are inferior to its counterparts of diagnostic CT. We propose to image the basic function unit of small/medium animal using diagnostic CT with an adaptor-and-holder assembly (AAHA). In data acquisition, the adaptor rotates in phase with x-ray source, but the animal holder spins at an angular velocity that is opposite to the direction of adaptor rotation. Consequently, the pose of the animal remains unchanged in data acquisition, which can reduce animal's visceral motion substantially. With such an AAHA device, the spatial resolution can be increased more than 2.5 times, while the temporal resolution can potentially be increased 2 times if super-short scan is exercised. In practice, the AAHA device may not align perfectly and rotate rigorously in phase with x-ray source. Hence, motioncompensated reconstruction is needed. With a preliminary evaluation conducted in the feasibility study, it is believed that, a diagnostic CT with the AAHA device may enable imaging small/medium animal at superior spatiotemporal and contrast resolution in comparison to micro-CT.

An energy-discrimination K-edge x-ray computed tomography (CT) system is useful for controlling the image contrast of a target region by selecting both the photon energy and the energy width. The CT system has an oscillation-type linear cadmium telluride (CdTe) detectror. CT is performed by repeated linear scans and rotations of an object. Penetrating x-ray photons from the object are detected by a CdTe detector, and event signals of x-ray photons are produced using charge-sensitive and shaping amplifiers. Both photon energy and energy width are selected out using a multichannel analyzer, and the number of photons is counted by a counter card. In energy-discrimination CT, the tube voltage and tube current were 80 kV and 20 μA, respectively, and the x-ray intensity was 1.92 μGy/s at a distance of 1.0 m from the source and a tube voltage of 80 kV. The energy-discrimination CT was carried out by selecting x-ray photon energies.

Novel geometrical designs of computed tomography (CT) scanners in combination with novel image reconstruction algorithms promise to reduce ionizing radiation exposure to the patient in CT scans. While the sampling density of the Field Of View (FOV) is retained, the image quality can even be increased in contrast to conventional CT scanners. In this study, we present first images obtained with a novel CT scanner that we developed in our working group. In this open CT system with irradiation within a fan beam, parallel Radon data are directly obtained for image reconstruction using the OPED (Orthogonal Polynomial Expansion on the Disk) algorithm. This algorithm uses Radon data directly, i.e., without any further data processing such as rebinning and interpolation. We experimentally test theoretical predictions for this system by quantifying image quality parameters in comparison with corresponding parameters that are derived from the images of a conventional scanner of the 3rd generation. The modulation transfer function (MTF) and noise power spectrum (NPS) are determined using a test phantom. The novel CT system quantitatively shows the same noise property as the conventional scanner. The resolution that is reached in the center of a reconstructed image is nearly identical for both scanner types. But we found that the resolution that is achieved in the novel CT system does not depend on the image position while the MTF of the conventional scanner decreases for radially outer regions of the image.

The simulation of imaging systems using Monte Carlo x-ray transport codes is a computationally intensive task. Typically, many days of computation are required to simulate a radiographic projection image and, as a consequence, the simulation of the hundreds of projections needed to perform a tomographic reconstruction may require an unaffordable amount of computing time. To speed up x-ray transport simulations, a MC code that can be executed in a graphics processing unit (GPU) was developed using the CUDATM programming model, an extension to the C language for the execution of general-purpose computations on NVIDIA's GPUs. The code implements the accurate photon interaction models from PENELOPE and takes full advantage of the GPU massively parallel architecture by simulating hundreds of particle tracks simultaneously. In this work we describe a new version of this code adapted to the simulation of computed tomography (CT) scans, and allowing the execution in parallel in multiple GPUs. An example simulation of a cardiac CT using a detailed voxelized anthropomorphic phantom is presented. A comparison of the simulation computational performance in one or multiple GPUs and in a CPU (Central Processing Unit), and a benchmark with a standard PENELOPE code, are provided. This study shows that low-cost GPU clusters are a good alternative to CPU clusters for Monte Carlo simulation of x-ray transport.

Accurately representing radiation dose delivered in MSCT is becoming a concern as the maximum beam width of some modern CT scanners tends to become wider than the 100 mm charge-collection length of the pencil ionization chamber generally used in CT dosimetry. We investigate two alternative methods of dose evaluation in CT scanners. We investigate two alternative approaches for better characterization of CT dose than conventional evaluation of CTDI₁₀₀. First, we simulate dose profiles and energy deposition in phantoms longer than the typically used 14-15 cm length right-circular cylinders. Second we explore the accuracy and practicality of applying mathematical convolution to a scatter kernel in order to generate dose profiles. A basic requirement for any newly designed phantom is that it be able to capture approximately the same dose as would an infinitely long cylinder, but yet be of a size and weight that a person could easily carry and position. Using the PENELOPE Monte Carlo package, we simulated dose profiles in cylindrical polymethyl methacrylate (PMMA) phantoms of 10, 16, 20, 24 and 32 cm diameter and 15, 30 and 300 cm length. Beam widths were varied from 1 cm to 60 cm. Lengths necessary to include within the dose integrals values associated with the scatter tails as well as with the primary radiation of the profile were then calculated as the full width at five percent of maximum dose. The resulting lengths suggest that to accommodate wide beam widths, phantoms longer than those currently used are necessary. The results also suggest that using a longer phantom is a relatively more accurate approach, while using mathematical convolution is simpler and more practical to implement than using the long phantoms designed according to direct Monte Carlo simulations.

While cone-beam CT using flat x-ray detectors has gained increased popularity in the past years, the 3D imaging quality is still limited by a large amount of scatter, low dynamic range, and small field of view of the detector. Especially for large objects, the high dynamic range of the projections is a common source for detector specific artifacts. In conventional CT, the application of beam shapers (or bowtie filters) to decrease the signal dynamic in the projections is quite common. In this paper we investigate the use of a beam shaper for cone-beam CT with an off-centered flat detector by means of Monte-Carlo (MC) simulations and test-bench experiments. The shift of the detector out of the central axis increases the field of view and allows the imaging of larger patients, but in turn leads to a very high dynamic signal range and poor scatter-to-primary ratios (SPR). The impact of a half bowtie filter on key performance parameters of the imaging chain is investigated with MC simulations. It is demonstrated that a beam shaper significantly improves the peak SPR especially for large patients and that the reshaping of the SPR has a dominant impact on the homogeneity of the reconstructed image. The use of beam shapers for CBCT requires a modified pre-processing chain that also accounts for secondary effects introduced by the beam modulation filter. Beside patient scatter correction, the inhomogeneous spectral hardening of the x-ray beam and scattered radiation from the beam shaper itself have to be corrected. A comparison of phantom scans with and without beam shaper after pre-processing demonstrates the potential of beam shapers for dose reduction and SNR improvement in flat detector cone-beam CT.

Registration and superimposition of images acquired from two different detectors is essential to dual-resolution cone beam CT. In this study we implemented and tested a method of integration, which is to register and superimpose the high resolution volume of interest (VOI) only images to the low resolution full field images. First, we acquired two images sets: One is low exposure low resolution full field images acquired with a low resolution detector; the other is high exposure high resolution volume of interest (VOI) images acquired with a high resolution detector and VOI mask. To locate the VOI positions in full field images, the third images set with VOI mask but without phantom was acquired with the low resolution detector. In the third images set, high contrast VOI boundaries were located and used to determine positions of the VOI in full field images. Then high resolution VOI images were superimposed with the full field images to generate integrated images set. Integrated images set was tested by subtraction from full field images set and then used to reconstruct images using regular FDK algorithm. In the reconstructed images, five Al wires (as small as 152 μm) can be clearly seen in the VOI.

Cone Beam Breast CT imaging (CBBCT) is a promising tool for diagnosis of breast tumors and calcifications. However, as the sizes of calcifications in early stages are very small, it is not easy to distinguish them from background tissues because of the relatively high noise level. Therefore, it is necessary to enhance the visualization of calcifications for accurate detection. In this work, the Papoulis-Gerchberg (PG) method was introduced and modified to improve calcification characterization. PG method is an iterative algorithm of signal extrapolation and has been demonstrated to be very effective in image restoration like super-resolution (SR) and inpainting. The projection images were zoomed by bicubic interpolation method, then the modified PG method were applied to improve the image quality. The reconstruction from processed projection images showed that this approach can effectively improve the image quality by improving the Modulation Transfer Function (MTF) with a limited increase in noise level. As a result, the detectability of calcifications was improved in CBBCT images.

Generally, Cone Beam CT imaging systems utilize polychromatic x-ray beams since they are easier to generate and their output power is larger than those from monochromatic beams. However, polychromatic beams have main drawbacks in medical imaging. For example, the relative high content of low energy photons induced two main phenomena that degrade the image quality. First, most of the low energy photons are absorbed within the soft tissue hardening the beam. This beam hardening deteriorates the image quality by decreasing the uniformity. Second, due to the polychromatic nature of the beam and beam hardening, the image is degraded by high contrast objects. In order to solve these problems, a filter was developed and implemented in a Flat Panel Detector-based cone Beam CT. First, the energy spectrum was acquired and computer simulations were done to evaluate and to develop the best suitable filter to harden the x-ray beam. Second, experiments were performed to evaluate the improvements on image quality based on the reduction of artifacts due to high contrast objects on two objects, namely a bone with titanium screws and a hand phantom. Lastly, the spatial resolution was evaluated to investigate the effects of this filter. The results demonstrate that the new developed filter improves the image quality in this area. The high contrast artifacts were also reduced as the images from the bone phantom with the metal implant illustrates. Evidently, the use of this filter to harden to beam has increased the image quality when metal implants are present.

Dynamic imaging with micro-CT often produces poorly-distributed sets of projections, and reconstructions of this data with filtered backprojection algorithms (FBP) may be affected by artifacts. Iterative reconstruction algorithms and total variation (TV) denoising are promising alternatives to FBP, but may require running times that are frustratingly long. This obstacle can be overcome by implementing reconstruction algorithms on graphics processing units (GPU). This paper presents an implementation of a family of iterative reconstruction algorithms with TV denoising on a GPU, and a series of tests to optimize and compare the ability of different algorithms to reduce artifacts. The mathematical and computational details of the implementation are explored. The performance, measured by the accuracy of the reconstruction versus the running time, is assessed in simulations with a virtual phantom and in an in vivo scan of a mouse. We conclude that the simultaneous algebraic reconstruction technique with TV minimization (SART-TV) is a time-effective reconstruction algorithm for producing reconstructions with fewer artifacts than FBP.

The Prior Image Constrained Compressed Sensing (PICCS) algorithm (Med. Phys. 35, pg. 660, 2008) has been applied to several computed tomography applications with both standard CT systems and flat-panel based systems designed for guiding interventional procedures and radiation therapy treatment delivery. The PICCS algorithm typically utilizes a prior image which is reconstructed via the standard Filtered Backprojection (FBP) reconstruction algorithm. The algorithm then iteratively solves for the image volume that matches the measured data, while simultaneously assuring the image is similar to the prior image. The PICCS algorithm has demonstrated utility in several applications including: improved temporal resolution reconstruction, 4D respiratory phase specific reconstructions for radiation therapy, and cardiac reconstruction from data acquired on an interventional C-arm. One disadvantage of the PICCS algorithm, just as other iterative algorithms, is the long computation times typically associated with reconstruction. In order for an algorithm to gain clinical acceptance reconstruction must be achievable in minutes rather than hours. In this work the PICCS algorithm has been implemented on the GPU in order to significantly reduce the reconstruction time of the PICCS algorithm. The Compute Unified Device Architecture (CUDA) was used in this implementation.

Ring artifacts often appear in flat-detector CT because of imperfect or defect detector elements or calibration. In high-spatial resolution CT images reducing such artifacts becomes a necessity. In this paper, we used the post-processing ring correction in polar coordinates (RCP)1 to eliminate the ring artifacts. The median filter is applied to the uncorrected images in polar coordinates and ring artifacts are extracted from the original images. The algorithm has a very high computational cost due to the time-expensive median filtering and coordinate transformation on CPUs. Graphics processing units (GPUs)ca n be seen as parallel co-processors with high computational power. All steps of the RCP algorithm were implemented with CUDA2(Compute Unified Device Architecture, NVIDIA). We introduced a new GPU-based branchless vectorized median (BVM)filter. 3, 4 This algorithm is based on minmax sorting and keeps track of a sorted array from which values are deleted and to which new values are inserted. For comparison purpose a modified pivot median filter5 on GPUs was presented, which compares a pivot element to all other values and recursively finds the median element. We evaluated the performance of the RCP method using 512 slices, each slice consisted of 512×512 pixels. This post-processing method efficiently reduces ring artifacts in the reconstructed images and improves image quality. Our CUDAbased RCP is up to 13.6 times faster than the optimized CPU-based (single core)r outine. Comparing our two GPU-based median filters showed a performance benefit by roughly 60% when switching from Pivot to BVM code. The main reason is that the BVM algorithm is branchless and makes use of data-level parallelism. The BVM method is better suited to the model of modern graphics processing. A multi-GPU solution showed that the performance scaled nearly linearly.

In this study, we demonstrated volume of interest (VOI) scanning technique in dual resolution cone beam CT (CBCT) breast imaging. A paraffin cylinder with a diameter of 130 mm was used to simulate breast. A wire phantom with a diameter of 15 mm was constructed as VOI. The phantom contains 8 vertically aluminum wires of various diameters surrounded by paraffin. The wire phantom was inserted into the breast phantom 45 mm away from the center. The phantoms were first scanned with a bench top experimental CBCT system at a low exposure level with the detector operated in a binning mode. Then a VOI mask was placed between the x-ray source and the phantoms. The phantoms were scanned again with high exposure level and the detector operated in the non-binning mode. The VOI mask was moved to follow the wire phantom during the whole CT scan to limit the exposures to cover the VOI only. The low resolution and high resolution images were then combined together for reconstruction with FDK algorithm. Visual review of the regular and dual resolution CBCT images shows that thinnest resolvable wire in the dual resolution CBCT images has a diameter of 152 μm. The thinnest resolvable wire in regular CBCT images has a diameter of 254 μm. The estimated dose to the phantom for dual resolution CBCT is 123% of that with regular CBCT at low exposure level. The dual resolution CBCT technique greatly enhances the CT image quality while still remains a low exposure level to the phantom.

This paper describes an initial investigation into means for producing lower-cost CT scanners for resource limited regions of the world. In regions such as sub-Saharan Africa, intermediate level medical facilities serving millions have no CT machines, and lack the imaging resources necessary to determine whether certain patients would benefit from being transferred to a hospital in a larger city for further diagnostic workup or treatment. Low-cost CT scanners would potentially be of immense help to the healthcare system in such regions. Such scanners would not produce state-of-theart image quality, but rather would be intended primarily for triaging purposes to determine the patients who would benefit from transfer to larger hospitals. The lower-cost scanner investigated here consists of a fixed digital radiography system and a rotating patient stage. This paper describes initial experiments to determine if such a configuration is feasible. Experiments were conducted using (1) x-ray image acquisition, a physical anthropomorphic chest phantom, and a flat-panel detector system, and (2) a computer-simulated XCAT chest phantom. Both the physical phantom and simulated phantom produced excellent image quality reconstructions when the phantom was perfectly aligned during acquisition, but artifacts were noted when the phantom was displaced to simulate patient motion. An algorithm was developed to correct for motion of the phantom and demonstrated success in correcting for 5-mm motion during 360-degree acquisition of images. These experiments demonstrated feasibility for this approach, but additional work is required to determine the exact limitations produced by patient motion.

Metallic implants are responsible for various artifacts in flat-detector computed tomography visible as streaks and dark areas in the reconstructed volumetric images. In this paper a novel method for a fast reduction of these metal artifacts is presented using a three-step correction procedure to approximate the missing parts of the raw data. In addition to image quality aspects, this paper deals with the problem of high correction latencies by proposing a reconstruction and correction framework, that utilizes the massive computational power of graphics processing units (GPUs). An initial volume is reconstructed, followed by a 3-dimensional metal voxel segmentation algorithm. These metal voxels allow us to identify metal-influenced detector elements by using a simplified geometric forward projection. Consequently, these areas are corrected using a 3D interpolation scheme in the raw data domain, followed by a second reconstruction. This volume is then segmented into three materials with respect to bone structures using a threshold-based algorithm. A forward projection of the obtained tissueclass model substitutes missing or corrupted attenuation values for each detector element affected by metal and is followed by a final reconstruction. The entire process including the initial reconstruction, takes less than a minute (5123 volume with 496 projections of size 1240x960) and offers significant improvements of image quality. The method was evaluated with data from two FD-CT C-arm systems (Artis Zee and Artis Zeego, Siemens Healthcare, Forchheim, Germany).

A high-speed motionless-gantry x-ray CT machine has been designed to allow for 3D images to be collected in real time. By using multiple, switched x-ray sources and fixed detector rings, the time consuming mechanical rotation of conventional CT machines can be removed. However, the nature of this design limits the possibility of detector collimation since each detector must now be able to record the energy of x-ray beams from a number of different directions. The lack of collimation has implications in the reconstructed image due to an increase in the number of scattered photons recorded. A Monte Carlo computer simulation of the x-ray machine has been developed, using the Geant4 software toolkit, to analyse the behaviour of both Rayleigh and Compton scattered photons when considering airport baggage and medical applications. Four different scattering objects were analysed based on 50kVp, 100kVp and 150kVp spectra for a tungsten target. Two suitcase objects, a body and a brain phantom were chosen as objects typical of airport baggage and medical CT. The results indicate that the level of scatter is negligible for a typical airport baggage application, since the majority of space in a suitcase consists of clothing, which has a low density. Scatter contributes to less than 1% of the image in all instances. However, due to the large amounts of water found in the human body, the level of scatter in the medical instances are significantly higher, reaching 37% when the body phantom is analysed at 50kVp.

Rotational angiography (RA) is widely used clinically to obtain 3D data. In many procedures, e.g., neurovascular interventions, the imaged field of view (FOV) is much larger than the region of interest (ROI), thereby subjecting the patient to unnecessary x-ray dose. To reduce the dose in these procedures, we have proposed placing an x-ray attenuating filter with an open aperture (ROI) in the x-ray beam (called filtered region of interest (FROI) RA. We have shown that this approach yields high quality data for centered objects of interest (OoIs). In this study, we investigate the noise behavior of the FROI approach for off-center OoIs. Using filter-specific attenuation and noise characteristics, simulated FROI projection images were generated. The intensities in the peripheral region were equalized, and the 3D data reconstructed. For each reconstructed voxel, the intersections with the full intensity beam (ROI) were determined for each projection, and noise properties were evaluated. Off-center OoIs intersect the high intensity beam in more than 60% of the projections (ROI having 40% FOV area), with intersection frequency increasing with increasing ROI area and OoI proximity to the central region. The noise increases with distance from the central region up to a factor of two. Integral dose reductions range between 40% and 85%, depending on ROI area and filter thickness. Substantial dose reductions (40-85%) are achieved with less than a factor of two increase in noise for OoIs peripheral to the central region, indicating the FROI approach might be an alternative for reducing dose during standard procedures.

Gating in small animal imaging can compensate for artifacts due to physiological motion. This paper presents a strategy for sampling and image reconstruction in the rodent lung using micro-CT. The approach involves rapid sampling of freebreathing mice without any additional hardware to detect respiratory motion. The projection images are analyzed postacquisition to derive a respiratory signal, which is used to provide weighting factors for each projection that favor a selected phase of the respiration (e.g. end-inspiration or end-expiration) for the reconstruction. Since the sampling cycle and the respiratory cycle are uncorrelated, the sets of projections corresponding to any of the selected respiratory phases do not have a regular angular distribution. This drastically affects the image quality of reconstructions based on simple filtered backprojection. To address this problem, we use an iterative reconstruction algorithm that combines the Simultaneous Algebraic Reconstruction Technique with Total Variation minimization (SART-TV). At each SART-TV iteration, backprojection is performed with a set of weighting factors that favor the desired respiratory phase. To reduce reconstruction time, the algorithm is implemented on a graphics processing unit. The performance of the proposed approach was investigated in simulations and in vivo scans of mice with primary lung cancers imaged with our in-house developed dual tube/detector micro-CT system. We note that if the ECG signal is acquired during sampling, the same approach could be used for phase-selective cardiac imaging.

The spectral sensitivity of quantum-counting detectors promises increased contrast-to-noise ratios and dualenergy capabilities for Computed Tomography (CT). In this article we quantify the benefits as well as the conceptual limitations of this technology under realistic clinical conditions. We present detailed simulations of a CT system with CdZnTe-based quantum-counting detector and compare to a conventional energy-integrating detector with Gd2O2S scintillator. Detector geometries and pixel layouts are adapted to specific requirements of clinical CT and its high-flux environment. The counting detector is realized as a two-threshold counter. An image-based method is used to adapt thresholds and data weights optimizing contrasts and image noise with respect to the typical spectra provided by modern high-power tungsten anode X-ray tubes. We consider the case of moderate X-ray fluxes and compare contrasts and image noise at same patient dose and image sharpness. We find that the spectral sensitivity of such a CT system offers dose reduction potentials of 31.5% (9.2%) maintaining Iodine-water contrast-to-noise ratios at 120kVp (80kVp). The improved contrast-to-noise ratios result mainly from improved contrasts and not from reduced image noise. The presence of fluorescence effects in the sensor material is the reason why image noise levels are not significantly reduced in comparison to energy-integrating systems. Dual-energy performance of quantum-counting single-source CT in terms of bone-Iodine separation is found to be somewhat below the level of today's dual-source CT devices with optimized pre-filtration of the X-ray beams.

Renal lesion detection and characterization using Computed Tomography is an important application in genitourinary radiology. Although in general the detection of renal lesions has been shown to be exceedingly accuratce, the detection of benign renal cysts is still problematic. Under certain circumstances, the attenuation values inside a cyst increase incorrectly with an increase in the iodine concentration in the surrounding soft tissue. This so called pseudoenhancement complicates the classification of cysts and creates severe difficulties to distinguish a benign nonenhancing lesion from an enhancing mass. In the present study, the standard procedure based on a single energy 120 kV mode is compared to three dual energy modes available on the Siemens Somatom Definition Flash scanner. In order to simulate the kidney and the lesions, several plastic rods were placed inside a small container filled with different iodine concentrations. This phantom is then positioned inside water tanks of different sizes. The rods simulating the lesions are made out of a special plastic with constant HU value throughout the relevant X-ray energy range. During the project, three important aspects have been discovered: 1) for normal situations, a 100/140 Sn kV mode on the Siemens Flash scanner is similar to the traditional single energy 120 kV mode. 2) For small patient sizes, all dual energy modes show a reduction of pseudoenhancement. 3) For larger patients, only the 100/140 Sn kV mode results in a reduction of pseudoenhancement. Both the 80/140 kV and the 80/140 Sn kV mode show a worse performance than the 120 kV single energy mode in a very large phantom size.

The clinical application of Gemstone Spectral Imaging^TM, a fast kV switching dual energy acquisition, is explored in the context of noninvasive kidney stone characterization. Utilizing projection-based material decomposition, effective atomic number and monochromatic images are generated for kidney stone characterization. Analytical and experimental measurements are reported and contrasted. Phantoms were constructed using stone specimens extracted from patients. This allowed for imaging of the different stone types under similar conditions. The stone specimens comprised of Uric Acid, Cystine, Struvite and Calcium-based compositions. Collectively, these stone types span an effective atomic number range of approximately 7 to 14. While Uric Acid and Calcium based stones are generally distinguishable in conventional CT, stone compositions like Cystine and Struvite are difficult to distinguish resulting in treatment uncertainty. Experimental phantom measurements, made under increasingly complex imaging conditions, illustrate the impact of various factors on measurement accuracy. Preliminary clinical studies are reported.

Photon counting detectors with energy discriminating capabilities offer the exciting prospect of dose efficient dual energy x-ray imaging. Several techniques have been proposed to form the energy-dependent detector measurements needed for dual energy. However, their performance depends on the limitations of the detectors, an important one being the detector's spectral response. Therefore, in this paper, we study the effect of a detector's spectral response on several dual energy techniques with photon counting detectors. We use a two-parameter model to characterize realistic spectral response functions, which exhibit a Gaussian photopeak trailed by a lower energy tail. The dual energy techniques compared all reduce the spectral data into two measurements using the following methods: (1) binning with an optimally chosen energy threshold; (2) a hybrid photon counting/energy integrating detector; (3) μ-weights, which have been shown to be optimal for ideal detectors. Their performances for different spectral responses are compared by evaluating the Cramer-Rao Lower Bound (CRLB) for estimating the unknown material thicknesses. We show that as the spectral response worsens, μ-weights rapidly deteriorate in performance while a hybrid detector's relative performance improves and consistently outperforms binning. Although the weighting functions are highly sensitive to a detector's spectral response, numerical optimization techniques can find two sets of weights whose performance is close to that of measuring the full detected spectrum.

In this paper we describe and evaluate an image-based spectral CT method. Its central formula expresses measured CT data as a spectral integration of the spectral attenuation coefficient multiplied by a LocalWeighting Function (LWF). The LWF represents the local energy weighting in the image domain, taking into account the system and reconstruction properties and the object self attenuation. A generalized image-based formulation of spectral CT algorithms is obtained, with no need for additional corrections of e.g. beam hardening. The iterative procedure called Local Spectral Reconstruction (LSR) yields both the mass attenuation coefficients of the object and a representation of the LWF. The quantitative accuracy and precision of the method is investigated in several applications, including beam hardening correction, attenuation correction for SPECT/CT and PET/CT and a direct identification of spectral attenuation functions using the LWF result is demonstrated. In all applications the ground truth of the objects is reproduced with a quantitative accuracy in the sub-percent to two percent range. An exponential convergence behavior of the iterative procedure is observed, with one to two iteration steps as a good compromise between quantitative accuracy and precision. We conclude that the method can be used to perform image-based spectral CT reconstructions with quantitative accuracy. Existing algorithms benefit from the intrinsic treatment of beam hardening and system properties. Novel algorithms are enabled to directly compare material model functions to spectral measurement data.

Dual energy computed tomography (CT) has been previously shown to be capable of quantifying iron concentration in a phantom model. In this work, a commercial three material decomposition algorithm was investigated with the aim of quantifying iron concentration in vivo with dual energy CT. Iron (III) nitrate solutions of five/seven different concentrations were placed in syringes of two different cross-sectional areas within anthropomorphic phantoms of three different sizes and scanned using various x-ray tube potentials and beam filtration levels. A commercial three material decomposition software package was used to measure iron concentration values in specified regions of interest. These data were used to assess the effects of tube potential, beam filtration, phantom size, and object size on the ability of dual energy CT to accurately quantify iron concentration. The object's cross sectional area (diameter of syringe containing the iron solution) affected the accuracy of the iron quantification, with measurements averaged over a larger region of interest having improved accuracy. In most cases, the greater spectral separation afforded by the tin filtration improved the accuracy of the iron quantification. Using the larger syringes (approximately 100 mm² cross sectional area) and small phantom size, dual energy CT measurements of the three highest iron concentrations (approximately 10 - 18 mg/ml) had a maximum percent difference from the known value of 21%.

In this study, the feasibility of differentiating uric acid from non-uric acid kidney stones in the presence of iodinated contrast material was evaluated using dual-energy CT (DECT). Iodine subtraction was accomplished with a commercial three material decomposition algorithm to create a virtual non-contrast (VNC) image set. VNC images were then used to segment stone regions from tissue background. The DE ratio of each stone was calculated using the CT images acquired at two different energies with DECT using the stone map generated from the VNC images. The performance of DE ratio-based stone differentiation was evaluated at five different iodine concentrations (21, 42, 63, 84 and 105 mg/ml). The DE ratio of stones in iodine solution was found larger than those obtained in non-iodine cases. This is mainly caused by the partial volume effect around the boundary between the stone and iodine solution. The overestimation of the DE ratio leads to substantial overlap between different stone types. To address the partial volume effect, an expectation-maximization (EM) approach was implemented to estimate the contribution of iodine and stone within each image pixel in their mixture area. The DE ratio of each stone was corrected to maximally remove the influence of iodine solutions. The separation of uric-acid and non-uric-acid stone was improved in the presence of iodine solution.

We model the effect of signal pile-up on the energy resolution of a photon counting silicon detector designed for high flux spectral CT with sub-millimeter pixel size. Various design parameters, such as bias voltage, lower threshold level for discarding of electronic noise and the entire electronic read out chain are modeled and realistic parameter settings are determined. We explicitly model the currents induced on the collection electrodes of a pixel and superimpose signals emanating from events in neighboring pixels, either due to charge sharing or signals induced during charge collection. Electronic noise is added to the pulse train before feeding it through a model of the read out electronics where the pulse height spectrum is saved to yield the detector energy response function. The main result of this study is that a lower threshold of 5 keV and a rather long time constant of the shaping filter (τ0 = 30 ns) are needed to discard induced pulses from events in neighboring pixels. These induction currents occur even if no charge is being deposited in the analyzed pixel from the event in the neighboring pixel. There is also only a limited gain in energy resolution by increasing the bias voltage to 1000 V from 600 V. We show that with these settings the resulting energy resolution, as measured by the FWHM/E of the photo peak, is 5% at 70 keV.

A detailed experimental and theoretical investigation of noise in both current mode and voltage mode amorphous silicon (a-Si) active pixel sensors (APS) has been performed. Both flicker (1/f) and thermal are considered in this study. The experimental result in this paper emphasizes the computation of the output noise variance. The theoretical analysis shows that the voltage mode APS has an advantage over the current mode APS in terms of the flicker noise due to the operation of the readout process. The experimental data are compared to the theoretical analysis and are in good agreement.

We propose a new indirect x-ray and gamma-ray detector which is comprised of a scintillating crystal coupled with an amorphous selenium (a-Se) metal-semiconductor-metal (MSM) photodetector. A lateral Frisch grid is embedded between the anode and the cathode to provide (1) unipolar charge sensing and (2) avalanche multiplication gain during hole transport inside the detection region. Unipolar charge sensing operation reduces the persistent photocurrent lag and increases the speed of the photodetector because most of the pixel charge is induced during carrier transport inside the detection region. Also, with proper biasing of the electrodes, we can create a high-field region between the lateral Frisch grid and the cathode for avalanche multiplication gain. Thus, we can convert the photodetector into a photomultiplier for higher signal-to-noise ratio and single photon-counting gamma-ray imaging. We present for the first time, a fabricated amorphous selenium lateral Frisch photodetector and present preliminary results of the measured photocurrents in response to a blue light emitting diode.

We are developing pixel-structured scintillators for the eventual purpose of high-resolution and high-sensitivity x-ray imaging applications. The pixel-structured scintillators were fabricated by filling Gd₂O₂S:Tb phosphor powder into the silicon micro-well arrays by using a simple sedimentation method. The micro-well arrays having a depth of 180 μm were fabricated by deep reactive ion etching of silicon wafers. To enhance the optical gain and the Swank noise factor, we applied reflectance at the inside wall surfaces. Two different inside-surface treatments were applied; 0.2-μm-thick titanium which has 70% reflectance and 1-μm-thick silicon dioxide which was grown by thermal oxidation. The imaging performance was evaluated in terms of modulation-transfer function (MTF), noise-power spectrum (NPS), and detective quantum efficiency (DQE). Compared with the commercial phosphor screen as a reference, much enhanced MTF results were measured. However, very low values of the system gain due to trapping of the generated optical photons at the wall surfaces give rise to the poorer DQE performance rather than that of the reference detector. The theoretical cascaded model analysis estimates much improved DQE performances with improved design parameters, such as higher reflectance of 90% at the wall surfaces.

Complementary metal-oxide-semiconductor (CMOS) active pixel sensors (APSs) with high electrical and optical performances are now being attractive for digital radiography (DR) and dental cone-beam computed tomography (CBCT). In this study, we report our prototype CMOS-based detectors capable of real-time imaging. The field-of-view of the detector is 12 × 14.4 cm. The detector employs a CsI:Tl scintillator as an x-ray-to-light converter. The electrical performance of the CMOS APS, such as readout noise and full-well capacity, was evaluated. The x-ray imaging characteristics of the detector were evaluated in terms of characteristic curve, pre-sampling modulation transfer function, noise power spectrum, detective quantum efficiency, and image lag. The overall performance of the detector is demonstrated with phantom images obtained for DR and CBCT applications. The detailed development description and measurement results are addressed. With the results, we suggest that the prototype CMOS-based detector has the potential for CBCT and real-time x-ray imaging applications.

We are developing a theoretical model to describe signal pulses from a detector-amplifier system operated in photoncounting mode. In the model, we include incomplete signal generation due to the charge trapping within a semiconductor detector as well as to the ballistic deficit caused by insufficient charge integration time. This model can be utilized for the characterization of detector material properties such as the mobility and the lifetime, as well as the optimization of operation conditions such as the applied bias voltages and the charge integration time. The model was experimentally verified with the measurement of charge collection efficiency of a planar cadmium zinc telluride detector with respect to the applied bias voltage and the charge integration time. We expect that the developed model will be helpful for the design of photon-counting detectors.

Active pixel sensor (APS) circuits are an alternate to passive pixel sensor (PPS) technology which, when integrated with a direct detection amorphous selenium (a-Se) photoconductor, can enable high performance, digital x-ray imaging applications such as real-time fluoroscopy due to their better signal-to-noise ratios at low dose. This paper presents experimental imaging results from a prototype 64×64 APS pixel array fabricated in a-Si technology. The prototype APS array is coated with a one millimeter thick layer of a-Se and the experimental results are evaluated using a standard radiography x-ray beam quality RQA5. The APS experimental results are compared with a standard real-time detector (FPD14) imaging array under the same x-ray beam conditions. In addition, we will theoretically examine the best achievable performance for our APS array fabricated in state-of-the-art a-Si technology and compare the results to state-of-the-art PPS panels for real-time fluoroscopy.

A simple benchtop apparatus has been built, to measure the x-ray imaging properties of fluorozirconate-based glassceramic x-ray storage phosphor materials. The MTF degradation due to stimulating light spreading in the plate is lower in comparison to optically turbid screens resulting in higher image MTF. In addition, the degree of transparency, or the amount of light scattering at the wavelength of the stimulating (laser) light is adjustable by means of the glass preparation process. The amount of stimulating exposure required for plate readout is generally higher than in previous systems, but well within the range of commercially available laser systems, for practical readout times. The effects of flare or unwanted readout due to back-reflection from the imaging plate is also less than in previous systems. A novel telecentric scanning system has been developed that is able to rapidly read out the latent image stored in the translucent imaging plates. This system features a reflective primary scan mirror to achieve telecentricity, optical correction for scan line bow, and the design should enable the construction of a relatively inexpensive scanner system for the translucent x-ray storage plates.

In indirect digital x-ray detectors, photodetectors such as hydrogenated amorphous silicon (a-Si:H) p-i-n photodetectors are used to convert the optical photons generated by the scintillating material to collectible electron-hole pairs. A problem that arises during the collection of the charges is that the mobility and lifetime of both types of carriers (electrons and holes) differ. In a-Si:H, the mobility of holes is much lower than that of electrons which leads to depth-dependent signal variations and causes the charge collection time to be extensive. It has been shown that the use of a Frisch grid can reduce the effect of the slower carriers in direct x-ray detectors. The Frisch grid is essentially a conducting grid that shields carriers from the collecting electrode until they are in close proximity. When the pixel electrodes are properly biased, the grid prevents the slow moving carriers (traveling away from the collecting electrode) from being collected and puts more weight on the fast moving carriers, thus allowing the total charge to be collected in less time. In this paper we investigate the use of a Frisch grid in a-Si:H p-i-n photodetectors for indirect x-ray detectors. Through simulations and theoretical analysis we determine the grid line sizes and positioning that will be most effective for practical p-i-n photodetector designs. In addition we compare the results of photodetectors with and without the grid to characterize the improvement achievable.

For a detector consisting of a phosphor screen and a photodiode array made by complementary metal-oxidesemiconductor (CMOS) process, we have experimentally re-investigated the long-term stability of the signal and noise characteristics as a function of the accumulated dose at the entrance surface of the detector in addition to the previous study [IEEE Trans. Nucl. Sci. 56(3) 1121 (2009)]. The irradiation and analysis were more systematically performed. We report the aging effect in image quality in terms of dark pixel signal, dynamic range, modulation-transfer function (MTF), and noise-power spectrum (NPS). Unlike the previous study, the electronic noise was dominantly increased with the total dose and the other statistical and structural noise sources were nearly independent on the cumulative dose. Similarly, the increase of dark pixel signal and the related noise gradually reduces the dynamic range as the total dose increases. While MTF was almost insensitive to the total dose, degradation in NPS was observed. Therefore, preprocessing without properly updated offset and gain images would underestimate the detective quantum efficiency when performing quality control of a detector in the field. Restoration of degraded dark signals due to aging is demonstrated by annealing the aged detector with thermal activation energy. This study provides a motivation that the periodic monitoring of the imagequality degradation is of great importance for the long-term and healthy use of digital x-ray imaging detectors.

Cadmium Zinc Telluride (CdZnTe or CZT) is a polycrystalline radiation detector that has been investigated over the years for a variety of applications including Constellation X-ray space mission [1] and direct-conversion medical imaging such as digital mammography [2]. Due to its high conversion gain and low electron-hole pair creation energy (~4.43 eV) [3], it has found use in high end, photon counting medical imaging applications including positron emission tomography (PET), computed tomography (CT) and single photon emission computed tomography (SPECT). However, its potential in low photon energy applications has not been fully explored. In this work, we explore the capacity of the CZT material to count low photon energies (6 keV - 20 keV). These energies are of direct relevance to applications in gamma ray breast brachytheraphy and mammography, X-ray protein crystallography, X-ray mammography and mammography tomosynthesis. We also present a design that integrates the CZT direct conversion detector with an inhouse fabricated amorphous silicon (a-Si:H) thin film transistor (TFT) passive pixel sensor (PPS) array. A CZT photoconductor (2 cm x 2 cm size, 5-mm-thick) prepared by the traveling heat method (THM) from Redlen^TM is characterized. The current-voltage characteristics reveal a resistivity of 3.3 x 1011 Ω•cm and a steady state dark current in the range of nA. Photocurrent transients under different biases and illumination pulses are studied to investigate photogeneration and the charge trapping process. It is found that charge trapping plays a more significant role in transient behavior at low biases and low frequency.

We have developed a novel direct conversion detector for digital radiography by using a fullerene (C₆₀)-doped polymer layer added on a thick amorphous selenium (a-Se) layer coupled to an amorphous silicon thin-film transistor (a-Si TFT) array. This detector exhibits considerable improvement in the lag characteristics and durability in high ambient temperatures. The C₆₀-doped polymer layer, which is directly and uniformly solution cast on the a-Se layer and followed by an inorganic electron-transporting layer, smoothly changes the electronic junction between the a-Se layer and the inorganic layer. It lubricates the emission of photocurrents from the a-Se photo-conversion layer and leads to the improved lag characteristics. Another merit of using a C₆₀-doped polymer is that it is stabile in high-temperature ambient conditions and is not degraded by humidity or a large amount of X-ray exposure. The polymer layer prevents the crystallization of a-Se, which otherwise occurs on exposure of a-Se to high temperature not only during the deposition of the inorganic layer or the metal electrode layer in the manufacturing process but also in actual use. A prototype detector, with a size of 17 in × 17 in and a pixel pitch of 150 μm, exhibited a good resolution; its DQE is approximately 48% at 1 cy/mm in 258 μC/kg (RQA5). This new development can simplify cooling apparatus and detector modules and also make a wide range of operational environments available. In addition, the improved lag characteristics make it possible to reduce the exposure intervals for static imaging, tomosynthesis, and other various exposure techniques.