Data-dependent Nonlinearity Analysis in CT Denoising CNNs

Recent years have seen the increasing application of deep learning methods in medical imaging formation, processing, and analysis. These methods take advantage of the flexibility of nonlinear neural network models to encode information and features in ways that can outperform conventional approaches. However, because of the nonlinear nature of this processing, images formed by neural networks have properties that are highly data-dependent and difficult to analyze. In particular, the generalizability and robustness of these approaches can be difficult to ascertain. In this work, we analyze a class of neural networks that use only piece-wise linear activation functions. This class of networks can be represented by locally linear systems where the linear properties are highly data-dependent - allowing, for example, estimation of noise in image output via standard propagation methods. We propose a nonlinearity index metric that quantifies the fidelity of a local linear approximation of trained models based on specific input data. We applied this analysis to three example CT denoising CNNs to analytically predict the noise properties in the output images. We found that the proposed nonlinearity metric highly correlates with the accuracy of noise predictions. The analysis proposed in this work provides theoretical understanding of the nonlinear behavior of neural networks and enables performance prediction and quantitation under certain conditions.

Fast CBCT Reconstruction using Convolutional Neural Networks for Arbitrary Robotic C-arm Orbits

Cone-beam CT (CBCT) with non-circular acquisition orbits has the potential to improve image quality, increase the field-of view, and facilitate minimal interference within an interventional imaging setting. Because time is of the essence in interventional imaging scenarios, rapid reconstruction methods are advantageous. Model-Based Iterative Reconstruction (MBIR) techniques implicitly handle arbitrary geometries; however, the computational burden for these approaches is particularly high. The aim of this work is to extend a previously proposed framework for fast reconstruction of non-circular CBCT trajectories. The pipeline combines a deconvolution operation on the backprojected measurements using an approximate, shift-invariant system response prior to processing with a Convolutional Neural Network (CNN). We trained and evaluated the CNN for this approach using 1800 randomized arbitrary orbits. Noisy projection data were formed from 1000 procedurally generated tetrahedral phantoms as well as anthropomorphic data in the form of 800 CT and CBCT images from the Lung Image Database Consortium Image Collection (LIDC). Using this proposed reconstruction pipeline, computation time was reduced by 90% as compared to MBIR with only minor differences in performance. Quantitative comparisons of nRMSE, FSIM and SSIM are reported. Performance was consistent for projection data simulated with acquisition orbits the network has not previously been trained on. These results suggest the potential for fast processing of arbitrary CBCT trajectory data with reconstruction times that are clinically relevant and applicable - facilitating the application of non-circular orbits in CT image-guided interventions and intraoperative imaging.

Three-dimensional printing of patient-specific lung phantoms for CT imaging: Emulating lung tissue with accurate attenuation profiles and textures

PURPOSE

Phantoms are a basic tool for assessing and verifying performance in CT research and clinical practice. Patient-based realistic lung phantoms accurately representing textures and densities are essential in developing and evaluating novel CT hardware and software. This study introduces PixelPrint, a 3D printing solution to create patient-based lung phantoms with accurate attenuation profiles and textures.

METHODS

PixelPrint, a software tool, was developed to convert patient digital imaging and communications in medicine (DICOM) images directly into FDM printer instructions (G-code). Density was modeled as the ratio of filament to voxel volume to emulate attenuation profiles for each voxel, with the filament ratio controlled through continuous modification of the printing speed. A calibration phantom was designed to determine the mapping between filament line width and Hounsfield units (HU) within the range of human lungs. For evaluation of PixelPrint, a phantom based on a single human lung slice was manufactured and scanned with the same CT scanner and protocol used for the patient scan. Density and geometrical accuracy between phantom and patient CT data were evaluated for various anatomical features in the lung.

RESULTS

For the calibration phantom, measured mean HU show a very high level of linear correlation with respect to the utilized filament line widths, (r > 0.999). Qualitatively, the CT image of the patient-based phantom closely resembles the original CT image both in texture and contrast levels (from -800 to 0 HU), with clearly visible vascular and parenchymal structures. Regions of interest comparing attenuation illustrated differences below 15 HU. Manual size measurements performed by an experienced thoracic radiologist reveal a high degree of geometrical correlation of details between identical patient and phantom features, with differences smaller than the intrinsic spatial resolution of the scans.

CONCLUSION

The present study demonstrates the feasibility of 3D-printed patient-based lung phantoms with accurate organ geometry, image texture, and attenuation profiles. PixelPrint will enable applications in the research and development of CT technology, including further development in radiomics.

Generative Adversarial Networks and Radiomics Supervision for Lung Lesion Synthesis

Realistic lesion generation is a useful tool for system evaluation and optimization. In this work, we investigate a data-driven approach for categorical lung lesion generation using public lung CT databases. We propose a generative adversarial network with a Wasserstein discrimination and gradient penalty to stabilize training. We further included conditional inputs such that the network can generate user-specified lesion categories. Novel to our network, we directly incorporated radiomic features in an intermediate supervision step to encourage similar textures between generated and real lesions. We evaluated the network using lung lesions from the Lung Image Database Consortium (LIDC) database. The lesions are divided into two categories: solid vs. non-solid. We performed quantitative evaluation of network performance base on four criteria: 1) overfitting in terms of structural and morphological similarity to the training data, 2) diversity of generated lesions in terms of similarity to other generated data, 3) similarity to real lesions in terms of distribution of example radiomics features, and 4) conditional consistency in terms of classification accuracy using a classifier trained on the training lesions. We imposed a quantitative threshold for similarity based on visual inspection. The percentage of non-solid and solid lesions that satisfy low overfitting and high diversity is 96.9% and 88.6% of non-solid and solid lesions respectively. The distribution of example radiomics features are similar in the generated and real lesions indicated by a low Kullback-Leibler divergence score. Classification accuracy for the generated lesions are comparable with that for the real lesions. The proposed network is a promising approach for data-driven generation of realistic lung lesions.

High-resolution model-based material decomposition in dual-layer flat-panel CBCT

PURPOSE

Spectral CT uses energy-dependent measurements that enable material discrimination in addition to reconstruction of structural information. Flat-panel detectors (FPDs) have been widely used in dedicated and interventional systems to deliver high spatial resolution, volumetric cone-beam CT (CBCT) in compact and OR-friendly designs. In this work, we derive a model-based method that facilitates high-resolution material decomposition in a spectral CBCT system equipped with a prototype dual-layer FPD. Through high-fidelity modeling of multilayer detector, we seek to avoid resolution loss that is present in more traditional processing and decomposition approaches.

METHOD

A physical model for spectral measurements in dual-layer flat-panel CBCT is developed including layer-dependent differences in system geometry, spectral sensitivities, and detector blur (e.g., due to varied scintillator thicknesses). This forward model is integrated into a model-based material decomposition (MBMD) method based on minimization of a penalized weighted least-squared (PWLS) objective function. The noise and resolution performance of this approach was compared with traditional projection-domain decomposition (PDD) and image-domain decomposition (IDD) approaches as well as one-step MBMD with lower-fidelity models that use approximated geometry, projection interpolation, or an idealized system geometry without system blur model. Physical studies using high-resolution three-dimensional (3D)-printed water-iodine phantoms were conducted to demonstrate the high-resolution imaging performance of the compared decomposition methods in iodine basis images and synthetic monoenergetic images.

RESULTS

Physical experiments demonstrate that the MBMD methods incorporating an accurate geometry model can yield higher spatial resolution iodine basis images and synthetic monoenergetic images than PDD and IDD results at the same noise level. MBMD with blur modeling can further improve the spatial-resolution compared with the decomposition results obtained with IDD, PDD, and MBMD methods with lower-fidelity models. Using the MBMD without or with blur model can increase the absolute modulation at 1.75 lp/mm by 10% and 22% compared with IDD at the same noise level.

CONCLUSION

The proposed model-based material decomposition method for a dual-layer flat-panel CBCT system has demonstrated an ability to extend high-resolution performance through sophisticated detector modeling including the layer-dependent blur. The proposed work has the potential to not only facilitate high-resolution spectral CT in interventional and dedicated CBCT systems, but may also provide the opportunity to evaluate different flat-panel design trade-offs including multilayer FPDs with mismatched geometries, scintillator thicknesses, and spectral sensitivities.

Direct reconstruction of anatomical change in low-dose lung nodule surveillance

Purpose: In sequential imaging studies, there exists rich information from past studies that can be used in prior-image-based reconstruction (PIBR) as a form of improved regularization to yield higher-quality images in subsequent studies. PIBR methods, such as reconstruction of difference (RoD), have demonstrated great improvements in the image quality of subsequent anatomy reconstruction even when CT data are acquired at very low-exposure settings. Approach: However, to effectively use information from past studies, two major elements are required: (1) registration, usually deformable, must be applied between the current and prior scans. Such registration is greatly complicated by potential ambiguity between patient motion and anatomical change-which is often the target of the followup study. (2) One must select regularization parameters for reliable and robust reconstruction of features. Results: We address these two major issues and apply a modified RoD framework to the clinical problem of lung nodule surveillance. Specifically, we develop a modified deformable registration approach that enforces a locally smooth/rigid registration around the change region and extend previous analytic expressions relating reconstructed contrast to the regularization parameter and other system dependencies for reliable representation of image features. We demonstrate the efficacy of this approach using a combination of realistic digital phantoms and clinical projection data. Performance is characterized as a function of the size of the locally smooth registration region of interest as well as x-ray exposure. Conclusions: This modified framework is effectively able to separate patient motion and anatomical change to directly highlight anatomical change in lung nodule surveillance.

A CT Denoising Neural Network with Image Properties Parameterization and Control

A wide range of dose reduction strategies for x-ray computed tomography (CT) have been investigated. Recently, denoising strategies based on machine learning have been widely applied, often with impressive results, and breaking free from traditional noise-resolution trade-offs. However, since typical machine learning strategies provide a single denoised image volume, there is no user-tunable control of a particular trade-off between noise reduction and image properties (biases) of the denoised image. This is in contrast to traditional filtering and model-based processing that permits tuning of parameters for a level of noise control appropriate for the specific diagnostic task. In this work, we propose a novel neural network that includes a spatial-resolution parameter as additional input permits explicit control of the noise-bias trade-off. Preliminary results show the ability to control image properties through such parameterization as well as the possibility to tune such parameters for increased detectability in task-based evaluation.

End-to-end Modeling for Predicting and Estimating Radiomics: Application to Gray Level Co-occurrence Matrices in CT

While radiomics models are finding increased use in computer-aided diagnostics and as imaging biomarkers for inference and discovery, their utility in computed tomography (CT) is limited by variability of the image properties produced by different CT scanners, imaging protocols, patient anatomy, and an increasingly diverse range of reconstruction and post-processing software. While these effects can be mitigated with careful data curation and standardization of protocols, this is impractical for diverse sources of image data. In this work, we propose to generalize traditional end-to-end imaging system models to include radiomics calculation as an explicit stage. Such a model permits both prediction of the undesirable variability of radiomics, but also forms a basis for inverting the process to estimate the true underlying radiomics. This framework has the potential to provide for standardization of radiomics across imaging conditions permitting more widespread application of radiomics models; larger, more diverse image databases; and improved diagnoses and inferences based on those standardized metrics. We apply this framework to a large class of popular radiomics based on the Gray Level Co-occurrence matrix under conditions of imaging system that are well describe by traditional linear systems approaches as well as nonlinear systems for which traditional analytic models do not apply.

Grating-based Spectral CT using Small Angle X-ray Beam Deflections

Interest in spectral CT for diagnostics and therapy evaluation has been growing. Acquisitions of data from distinct energy spectra provide, among other advantages, quantitative density estimations for multiple materials. We introduce a novel spectral CT concept that includes a fine-pitch grating for prefiltration of the x-ray beam. The attenuation behavior of this grating changes significantly if x-rays are slightly angled in relation to the grating structures. To apply such an angle (i.e. switch between the different filtrations) we propose a fast, controllable, and precise solution by moving the focal spot of the x-ray tube. In this work, we performed preliminary evaluations with a grating prototype on a CT test bench. Our results include x-ray spectrometer measurements that reveal diverse and controllable spectral shaping between 4° and 6° for a specific grating design. Additional experiments with a contrast agent phantom illustrated the capability to decompose clinically relevant iodine concentrations (5, 10, 20, and 50mg/mL) - demonstrating the feasibility of the grating-based approach. Ongoing and future studies will investigate the potential of this novel concept as a relatively simple upgrade to standard energy-integrating CT.

Metal-Tolerant Noncircular Orbit Design and Implementation on Robotic C-Arm Systems

Metal artifacts are a major confounding factor for image quality in CT, especially in image-guided surgery scenarios where surgical tools and implants frequently occur in the field-of-view. Traditional metal artifact correction methods typically use algorithmic solutions to interpolate over the highly attenuated projection measurements where metal is present but cannot recover the missing information obstructed by the metal. In this work, we treat metal artifacts as a missing data problem and employ noncircular orbits to maximize data completeness in the presence of metal. We first implement a local data completeness metric based on Tuy's condition as the percentage of great circles sampled by a particular orbit and accounted for the presence of metal by discounting any rays that pass through metal. We then compute the metric over many locations and many possible metal locations to reflect data completeness for arbitrary metal placements within a volume of interest. We used this metric to evaluate the effectiveness of sinusoidal orbits of different magnitudes and frequencies in metal artifact reduction. We also evaluated noncircular orbits in two imaging systems for phantoms with different metal objects and metal arrangements. Among a circular, tilted circular, and a sinusoidal orbit of two cycles per rotation, the latter is shown to most effectively remove metal artifacts. The noncircular orbit not only reduce the extent of streaks, but allows better visualization of spatial frequencies that cannot be recovered by metal artifact correction algorithms. These results illustrate the potential of relatively simple noncircular orbits to be robust against metal implants which ordinarily present significant challenges in interventional imaging.

Perturbation Response of Model-based Material Decomposition with Edge-Preserving Penalties

Spectral CT permits material discrimination beyond the structural information in conventional single-energy CT. Model-based material decomposition facilitates direct estimation of material density from spectral measurements, incorporating a general forward model for arbitrary spectral CT system, a statistical model of spectral CT measurements, and flexible regularization schemes. Such one-step approaches are promising for superior image quality, but the relationship between regularization parameters, imaging conditions, and reconstructed image properties is complicated. More specifically, the estimator is inherently nonlinear and may include additional nonlinearities like edge-preserving regularization, making image quality metrics intended for linear system evaluation difficult to apply. In this work, we seek approaches to quantify the image properties of this inherently nonlinear process through an investigation of perturbation response - the generalized system response to a local perturbation of arbitrary shape, location, and contrast. Such responses include cross-talk between material density channels, and we investigate the application of this metric in a sample spectral CT system. Inspired by the prior work under assumptions of local linearity and shift-invariant we also propose a prediction framework for perturbation response using a perceptron neural network. The proposed prediction framework offers an alternative to exhaustive evaluation and is a potential tool that can be used to prospectively choose optimal regularization parameters based on imaging conditions and diagnostic task.

Combining Spectral CT Acquisition Methods for High-Sensitivity Material Decomposition

Quantitative estimation of contrast agent concentration is made possible by spectral CT and material decomposition. There are several approaches to modulate the sensitivity of the imaging system to obtain the different spectral channels required for decomposition. Spectral CT technologies that enable this varied sensitivity include source kV-switching, dual-layer detectors, and source-side filtering (e.g., tiled spatial-spectral filters). In this work, we use an advanced physical model to simulate these three spectral CT strategies as well as hybrid acquisitions using all combinations of two or three strategies. We apply model-based material decomposition to a water-iodine phantom with iodine concentrations from 0.1 to 5.0 mg/mL. We present bias-noise plots for the different combinations of spectral techniques and show that combined approaches permit diversity in spectral sensitivity and improve low concentration imaging performance relative to the those strategies applied individually. Better ability to estimate low concentrations of contrast agent has the potential to reduce risks associated with contrast administration (by lowering dosage) or to extend imaging applications into targets with much lower uptake.

Performance Assessment of Texture Reproduction in High-Resolution CT

Assessment of computed tomography (CT) images can be complex due to a number of dependencies that affect system performance. In particular, it is well-known that noise in CT is object-dependent. Such object-dependence can be more pronounced and extend to resolution and image textures with the increasing adoption of model-based reconstruction and processing with machine learning methods. Moreover, such processing is often inherently nonlinear complicating assessments with simple measures of spatial resolution, etc. Similarly, recent advances in CT system design have attempted to improve fine resolution details - e.g., with newer detectors, smaller focal spots, etc. Recognizing these trends, there is a greater need for imaging assessment that are considering specific features of interest that can be placed within an anthropomorphic phantom for realistic emulation and evaluation. In this work, we devise a methodology for 3D-printing phantom inserts using procedural texture generation for evaluation of performance of high-resolution CT systems. Accurate representations of texture have previously been a hindrance to adoption of processing methods like model-based reconstruction, and texture serves as an important diagnostic feature (e.g. heterogeneity of lesions is a marker for malignancy). We consider the ability of different systems to reproduce various textures (as a function of the intrinsic feature sizes of the texture), comparing microCT, cone-beam CT, and diagnostic CT using normal- and high-resolution modes. We expect that this general methodology will provide a pathway for repeatable and robust assessments of different imaging systems and processing methods.

Non-circular CT orbit design for elimination of metal artifacts

Metal artifacts are a well-known problem in computed tomography - particularly in interventional imaging where surgical tools and hardware are often found in the field-of-view. An increasing number of interventional imaging systems are capable of non-circular orbits providing one potential avenue to avoid metal artifacts entirely by careful design of the orbital trajectory. In this work, we propose a general design methodology to find complete data solution by applying Tuy's condition for data completeness. That is, because metal implants effectively cause missing data in projections, we propose to find orbital designs that will not have missing data based on arbitrary placement of metal within the imaging field-of-view. We present the design process for these missing-data-free orbits and evaluate the orbital designs in simulation experiments. The resulting orbits are highly robust to metal objects and show greatly improved visualization of features that are ordinarily obscured.

Model-based Material Decomposition with System Blur Modeling

In this work, we present a novel model-based material decomposition (MBMD) approach for x-ray CT that includes system blur in the measurement model. Such processing has the potential to extend spatial resolution in material density estimates - particularly in systems where different spectral channels exhibit different spatial resolutions. We illustrate this new approach for a dual-layer detector x-ray CT and compare MBMD algorithms with and without blur in the reconstruction forward model. Both qualitative and quantitative comparisons of performance with and without blur modeling are reported. We find that blur modeling yields images with better recovery of high-resolution structures in an investigation of reconstructed line pairs as well as lower cross-talk bias between material bases that is ordinarily found due to mismatches in spatial resolution between spectral channels. The extended spatial resolution of the material decompositions has potential application in a range of high-resolution clinical tasks and spectral CT systems where spectral channels exhibit different spatial resolutions.

Prospective Prediction and Control of Image Properties in Model-based Material Decomposition for Spectral CT

Model-based material decomposition (MBMD) directly estimates the material densities from the spectral CT data and has found opportunities for dose reduction via physical and statistical modeling and advanced regularization. However, image properties such as spatial resolution, noise, and cross-basis response in the context of material decomposition are dependent on regularization, and high-dimensional exhaustive sweeping of regularization parameters is suboptimal. In this work, we proposed a set of prediction tools for generalized local impulse response (LIR) that characterizes both in-basis spatial resolution and cross-basis response, and noise correlation prospectively. The accuracy of noise predictor was validated in a simulation study, comparing predicted and measured in- and cross-basis noise correlations. Employing these predictors, we composed a specialized regularization for cross-talk reduction and showed that such prediction tools are promising for task-based optimization in spectral CT applications.

An Investigation of Slot-scanning for Mammography and Breast CT

Mammography and breast CT are important tools for breast cancer screening and diagnosis. Current implementations are limited by scattered radiation and/or spatial resolution. In this work, we propose and develop a slot scan-based system to be used in both mammography and CT mode that can limit scatter and collect sparse CT data for improved image quality at low radiation exposures. Monte Carlo simulations of an anthropomorphic breast phantom show a factor of 10 reduction in scattering amplitude with our slot scan-based system compared to that of a full-field detector mammography system (area mode). Similarly, slot-scan improved the MTF (particularly the low-frequency response) compared to an area detector. Investigation of sparse CT sampling with doubly sparse acquisition data return better quality reconstruction, for which our slot-scanning system is capable, over angle-only projection. Thus, a system with the combined ability for slot-scanning mammography and slot-scanning breast CT has the potential to deliver improved dose-efficient imaging performance and become viable breast cancer screening and diagnostic tools.

Volume-of-interest imaging with dynamic fluence modulation using multiple aperture devices

Volume-of-interest (VOI) imaging is a strategy in computed tomography (CT) that restricts x-ray fluence to particular anatomical targets via dynamic beam modulation. This permits dose reduction while retaining image quality within the VOI. VOI-CT implementation has been challenged, in part, by a lack of hardware solutions for tailoring the incident fluence to the patient and anatomical site, as well as difficulties involving interior tomography reconstruction of truncated projection data. We propose a general VOI-CT imaging framework using multiple aperture devices (MADs), an emerging beam filtration scheme based on two binary x-ray filters. Location of the VOI is prescribed using two scout views at anterior-posterior (AP) and lateral perspectives. Based on a calibration of achievable fluence field patterns, MAD motion trajectories were designed using an optimization objective that seeks to maximize the relative fluence in the VOI subject to minimum fluence constraints. A modified penalized-likelihood method is developed for reconstruction of heavily truncated data using the full-field scout views to help solve the interior tomography problem. Physical experiments were conducted to show the feasibility of noncentered and elliptical VOI in two applications-spine and lung imaging. Improved dose utilization and retained image quality are validated with respect to standard full-field protocols. We observe that the contrast-to-noise ratio (CNR) is 40% higher compared with low-dose full-field scans at the same dose. The total dose reduction is 50% for equivalent image quality (CNR) within the VOI.

Comparison of deep learning and human observer performance for detection and characterization of simulated lesions

Detection and characterization of abnormalities in clinical imaging are of utmost importance for patient diagnosis and treatment. We present a comparison of convolutional neural network (CNN) and human observer performance on a simulated lesion detection and characterization task. We apply both conventional performance metrics, including accuracy and nonconventional metrics such as lift charts to perform qualitative and quantitative comparisons of each type of observer. It is determined that the CNN generally outperforms the human observers, particularly at high noise levels. However, high noise correlation reduces the relative performance of the CNN, and human observer performance is comparable to CNN under these conditions. These findings extend into the field of diagnostic radiology, where the adoption of deep learning is starting to become widespread. Consideration of the applications for which deep learning is most effective is of critical importance to this development.

Performance Analysis for Nonlinear Tomographic Data Processing

Image quality analysis of nonlinear algorithms is challenging due to numerous dependencies on the imaging system, algorithmic parameters, object, and stimulus. In particular, traditional notions of linearity and local linearity are of limited utility when the system response is dependent on the stimulus itself. In this work, we analyze the performance of nonlinear systems using perturbation response - the difference between the mean output with and without a stimulus, and introduce a new metric to examine variation of the responses in individual images. We applied the analysis to four algorithms with different degrees of nonlinearity for a spherical stimulus of varying contrast. For model-based reconstruction methods [penalized-likelihood (PL) reconstruction with a quadratic penalty and a Huber penalty], perturbation response analysis reaffirmed known trends in terms of object- and location-dependence. For a CNN denoising network, the response exhibits highly nonlinear behavior as the contrast increases - from the stimulus completely disappearing, to appearing at the right contrast but smaller in size, to being fully admitted by the algorithm. Furthermore, the variation metric for PL reconstruction with a Huber penalty and the CNN network reveals high variation at the edge of the stimulus, i.e., perturbation response computed from the mean images is a smoothed version of individual responses due to "jitter" in edges. This behavior suggests that the mean response alone may not be representative of performance in individual images and image quality metrics traditionally defined based on the mean response may be inappropriate for certain nonlinear algorithms. This work demonstrates the potential utility of perturbation response and response variation in the analysis and optimization of nonlinear imaging algorithms.

Dynamic fluence field modulation in computed tomography using multiple aperture devices

A novel beam filter consisting of multiple aperture devices (MADs) has been developed for dynamic fluence field modulation (FFM) in CT. Each MAD achieves spatial modulation of x-ray through fine-scale, highly attenuating tungsten bars of varying widths and spacings. Moiré patterns produced by relative motions between two MADs provide versatile classes of modulation profiles. The dual-MAD filter can be designed to achieve specific classes of target profiles. The designed filter was manufactured through a laser-sintering process and integrated to an experimental imaging system that enables linear actuation of the MADs. Dynamic FFM was achieved through a combination of beam shape modulation (by relative MAD motion) and amplitude modulation (by view-dependent mAs). To correct for gains associated with the MADs, we developed an algorithm to account for possible focal spot changes during/between scans and spectral effects introduced by the MADs. We performed FFM designs for phantoms following two imaging objectives: (1) to achieve minimum mean variance in filtered backprojection (FBP) reconstruction, and (2) to flatten the fluence behind the phantom. Comparisons with conventional FFM strategies involving a static bowtie and pulse width modulation were performed. The dual-MAD filter produced modulation profiles closely matched with the design target, providing varying beam widths not achievable by the static bowtie. The entire range of modulation profiles was achieved by 0.373 mm of MAD displacement. The correction algorithm effectively alleviated ring artifacts as a result of MADs while preserving phantom details such as wires and tissue boundaries. Dynamic FFM enabled by the MADs were effective in achieving the imaging objectives and demonstrated superior FFM capabilities compared to the static bowtie. In an ellipse phantom, the FFM of objective 1 achieved the lowest mean variance in all cases investigated. The FFM of objective 2 produce nearly isotropic local noise power spectrum and homogeneous noise magnitude. The dual-MAD filter provides an effective tool for fluence control in CT to overcome limitations of conventional static bowties and to further enable patient-specific FFM studies for a wide range of dose and image quality objectives.

Generalized Prediction Framework for Reconstructed Image Properties using Neural Networks

Model-based reconstruction (MBR) algorithms in CT have demonstrated superior dose-image quality tradeoffs compared to traditional analytical methods. However, the nonlinear and data-dependent nature of these algorithms pose significant challenges for performance evaluation and parameter optimization. To address these challenges, this work presents an analysis framework for quantitative and predictive modeling of image properties in general nonlinear MBR algorithms. We propose to characterize the reconstructed appearance of arbitrary stimuli by the generalized system response function that accounts for dependence on the imaging conditions, reconstruction parameters, object, and the stimulus itself (size, contrast, location). We estimate this nonlinear function using a multilayer perceptron neural network by providing input and output pairs that samples the range of imaging parameters of interest. The feasibility of this approach was demonstrated for predicting the appearance of a spiculated lesion reconstructed by a penalized-likelihood objective with a Huber penalty in a physical phantom as a function of its location and reconstruction parameters β and δ. The generalized system response functions predicted from the trained neural network show good agreement with those computed from mean reconstructions, proving the ability of the framework in mapping out the nonlinear function for combinations of imaging parameters not present in the training data. We demonstrated utility of the framework to achieve desirable (e.g., non-blocky) lesion appearance in arbitrary locations in the phantom without the need for performing actual reconstructions. The proposed prediction framework permits efficient and quantifiable performance evaluations to provide robust control and understanding of image properties for general classes of nonlinear MBR algorithms.

Task-driven source-detector trajectories in cone-beam computed tomography: I. Theory and methods

We develop a mathematical framework for the design of orbital trajectories that are optimal to a particular imaging task (or tasks) in advanced cone-beam computed tomography systems that have the capability of general source-detector positioning. The framework allows various parameterizations of the orbit as well as constraints based on imaging system capabilities. To accommodate nonstandard system geometries, a model-based iterative reconstruction method is applied. Such algorithms generally complicate the assessment and prediction of reconstructed image properties; however, we leverage efficient implementations of analytical predictors of local noise and spatial resolution that incorporate dependencies of the reconstruction algorithm on patient anatomy, x-ray technique, and geometry. These image property predictors serve as inputs to a task-based performance metric defined by detectability index, which is optimized with respect to the orbital parameters of data acquisition. We investigate the framework of the task-driven trajectory design in several examples to examine the dependence of optimal source-detector trajectories on the imaging task (or tasks), including location and spatial-frequency dependence. A variety of multitask objectives are also investigated, and the advantages to imaging performance are quantified in simulation studies.

Volume-of-interest Imaging Using Multiple Aperture Devices

Volume-of-interest (VOI) imaging is a promising strategy for dose reduction in computed tomography (CT) while retaining image quality. However, implementation of VOI-CT has been challenged by the lack of adequate hardware and the interior tomography reconstruction problem. Multiple aperture devices (MAD) are a novel filtration scheme that can achieve x-ray fluence field modulation in a compact design with small translations. In this work, we propose a general approach for VOI imaging using MADs. MAD trajectories are designed to dynamically tailor the fluence for prescribed VOI. A penalized-likelihood reconstruction algorithm is proposed for fully truncated projections extended with scout views. Physical experiments were conducted to verify the feasibility for non-centered elliptic VOIs. Image quality and dose were estimated and compared with standard fullfield protocols. The ability of MAD-based VOI imaging to retain high image quality while significantly decreasing the total dose is demonstrated, suggesting the potential for dose reduction in clinical CT applications.

Regularization Analysis and Design for Prior-Image-Based X-Ray CT Reconstruction

Prior-image-based reconstruction (PIBR) methods have demonstrated great potential for radiation dose reduction in computed tomography applications. PIBR methods take advantage of shared anatomical information between sequential scans by incorporating a patient-specific prior image into the reconstruction objective function, often as a form of regularization. However, one major challenge with PIBR methods is how to optimally determine the prior image regularization strength which balances anatomical information from the prior image with data fitting to the current measurements. Too little prior information yields limited improvements over traditional model-based iterative reconstruction, while too much prior information can force anatomical features from the prior image not supported by the measurement data, concealing true anatomical changes. In this paper, we develop quantitative measures of the bias associated with PIBR. This bias exhibits as a fractional reconstructed contrast of the difference between the prior image and current anatomy, which is quite different from traditional reconstruction biases that are typically quantified in terms of spatial resolution or artifacts. We have derived an analytical relationship between the PIBR bias and prior image regularization strength and illustrated how this relationship can be used as a predictive tool to prospectively determine prior image regularization strength to admit specific kinds of anatomical change in the reconstruction. Because bias is dependent on local statistics, we further generalized shift-variant prior image penalties that permit uniform (shift invariant) admission of anatomical changes across the imaging field of view. We validated the mathematical framework in phantom studies and compared bias predictions with estimates based on brute force exhaustive evaluation using numerous iterative reconstructions across regularization values. The experimental results demonstrate that the proposed analytical approach can predict the bias-regularization relationship accurately, allowing for prospective determination of the prior image regularization strength in PIBR. Thus, the proposed approach provides an important tool for controlling image quality of PIBR methods in a reliable, robust, and efficient fashion.