To improve temporal resolution in prospectively gated axial cardiac CT scans, short scan (half-scan, partial scan) is used for image reconstruction. While some vendors offer scanners with 16cm collimation, capable of collecting entire heart data in a single rotation, the majority of routine cardiac scans are still done with 4cm collimation. In case of a prospective axial cardiac scanning, four or more axial acquisitions are performed at staggered patient table positions to cover the entire heart. At each acquisition, raw data is collected at the prescribed phase of cardiac R-R interval with the range of the x-ray source angles covering one or less than one rotation. If this angle range is greater than what is required for a short scan reconstruction, it allows some room for optimizing the reconstruction phase. Often, such optimization is done by manually reviewing images at slightly different reconstruction phase angles, and selecting the images with the least pronounced motion artifacts. Considering there are at least four acquisitions for each prospective cardiac scan, this may become a tedious time-consuming process. This paper proposes an automated process to select the best short-scan view range within full rotation acquisitions that minimizes motion artifacts at each table position. The proposed method was tested with a motion phantom which was connected to an ECG simulator and clinical cardiac data. Results show that the proposed method reliably provides reduction of motion artifact in reconstructed images.
In this work we revisit TV filter and propose an improved version that is tailored to diagnostic CT purposes. We revise TV cost function, which results in symmetric gradient function that leads to more natural noise texture. We apply a multi-scale approach to resolve noise grain issue in CT images. We examine noise texture, granularity, and loss of low contrast in the test images. We also discuss potential acceleration by Nesterov and Conjugate Gradient methods.
Motion estimation is a very important method for improving image quality by compensating the cardiac motion at the best phase reconstructed. We tackle the cardiac motion estimation problem using an image registration approach. We compare the performance of three gradient-based registration methods on clinical data. In addition to simple gradient descent, we test the Nesterov accelerated descent and conjugate gradient algorithms. The results show that accelerated gradient methods provide significant speedup over conventional gradient descent with no loss of image quality.
A new motion estimation and compensation method for cardiac computed tomography (CT) was developed. By
combining two motion estimation (ME) approaches the proposed method estimates the local and global cardiac motion
and then preforms motion compensated reconstruction. The combined motion estimation method has two parts: one is
the local motion estimation, which estimates the coronary artery motion by using coronary artery tree tracking and
registration; the other one is the global motion estimation, which estimates the entire cardiac motion estimation by image
registration. The final cardiac motion is the linear combination of the coronary artery motion and entire cardiac motion
the. We use the backproject-then-warp method proposed by Pack et al. to perform motion compensation reconstruction
(MCR). The proposed method was evaluated with 5 patient data and improvements in sharpness of both coronary
arteries and heart chamber boundaries were obtained.
CT image quality is affected by various artifacts including noise. Among these artifacts of different causes, noisy data
due to photon starvation should be contained in early processing stage to better mitigate other artifacts as they can cause
severe streaks and noise in reconstructed CT image. For low dose imaging, it is critical to use effective processing
method to handle the photon starved data in order to obtain required image quality with desired resolution, texture, low
contrast detectability. In this paper, two promising projection domain noise reduction methods are proposed. They are
derived from (1) the noise model that connects the noise behaviors in count and attenuation; (2) predicted noise
reduction from a finite impulse response (FIR) filter; (3) two pre-determined noise reduction requirements (noise
equalization and electronic noise suppression). Both methods showed significant streaks and noise reduction in tested
cases while reasonably maintaining the resolution of the images.
A typical iterative CT reconstruction using SART involves ray-driven forward projection and voxel-driven back-
ward projection. Bilinear interpolation is usually applied on image data for forward projection, and linear
interpolation is usually applied on projection data for backward projection, when both data are represented
using discrete samples in 2D fan-beam geometry. The applied interpolations, however, may affect the spatial
resolution, bias and noise properties of the reconstruction. A basis function (such as blob and spline) is therefore
applied to formulate a continuous model for the image data to reduce bias. In this paper we propose to apply
the blob representation on the projection data and explore its effectiveness. In this way we use continuous model
for the data of projection difference during backward projection, and we avoid the linear interpolation in this
process. Experimental results show that the proposed scheme is able to provide higher spatial resolution than
linear interpolation, while introducing more local variations in the reconstruction. However, the introduced local
variations may be reduced with the combination of total variation (TV) minimization. The proposed scheme is
therefore able to provide improved spatial resolution while keeping low local variations in reconstructions.
Advantages of iterative reconstruction (IR) algorithms over standard filtered backprojection (FBP) algorithms include
improved resolution and better noise performance, and many IR algorithms have been described in the literature. More
recently model-based IR algorithms (MBIR) have been developed, which incorporate accurate system models into IR,
resulting in better image quality than IR algorithms without a system model. This work investigates the resolution
improvement achieved when a system optics model (SOM) has been included in a standard OS-SART algorithm.
Three OS-SART algorithms have been compared: (1) "Pencil beam"
(IR-P) with no system optics; (2) system optics
included in forward projection (IR-SOM-FP), and (3) system optics included in both forward and backprojection
(IRSOM-FPBP). Simulated reconstructions of a 0.2 mm bead show that IR-SOM-FPBP produced a FWHM resolution of
0.41 mm, considerably better than FBPJ (0.87 mm), IR-P (0.63 mm), and IR-SOM-FP (0.59 mm).
Electronic noise becomes a major source of signal degradation in
low-dose clinical computed tomography (CT). In
current clinical scanners based on energy integrating x-ray detectors, electronic noise from the readout circuits adds a
noise of constant variance, which is negligible at high counts but can be significant at low count levels. On the other
hand, in a photon counting detector (PCD) with pulse height discrimination capability, electronic noise has little to no
impact on the measured signal. PCDs are known for their abilities to provide useful spectral information. In this work,
we investigate this dose reduction to improve low-dose single-energy CT. We perform low-dose single-energy CT
simulations using both energy integrating and photon counting detectors, and compare results with both analytical and
iterative reconstructions (IR). The results demonstrate the dose reduction potential of PCDs in conventional low-dose
single-energy CT examinations, when spectral information is not required.
We propose a method for material separation using dual energy data. Our method is suitable to separation of three or
more materials. In this work we describe our method and show results of numerical simulation and with real dual-energy
data of a head phantom. The proposed method of constructing the material separation map consists of the following
steps: Data-domain dual energy decomposition - Vector plot - Density plot - Clustering - Color assignment. Density
plots are introduced to allow automatic cluster separation. We use special image processing methods, including Gaussian
decomposition, to improve the accuracy of material separation. We also propose using the HSL color model for better
visualization and to bring a new dimension in material separation display. We study applications of bone removal and
virtual contrast removal. Evaluation shows improved accuracy compared to standard methods.
The spatial characteristics of noise in X-ray CT can influence object detectability. Now, as three dimensional image
reformations become more clinically common, it has become vital to understand the structure of noise in the x, y, and z
directions independently. The purpose of this paper is to study noise structure in the radial direction and tangential
direction, under varying conditions, including a wedge filter and acquisition techniques (i.e. half scan vs full scan).
Because the effect of the reconstruction algorithm on an image is highly dependent on spatial location within the field of
view, the effect of off-center vs centered positioning in each direction is also examined. The noise spatial frequency
distribution was investigated via calculation of the noise power spectrum (NPS) through Fourier methods on simulated
water images. As expected, noise structure at center was equivalent in both the radial and tangential directions.
Towards the periphery, overall noise power was muted. However, in the tangential direction high frequency noise power
was preserved more than twice as much as in the radial direction. Towards the periphery noise becomes more low-frequency
in the radial direction, while in the tangential direction it becomes more high-frequency. The half scan
increased both noise magnitude and low-mid spatial correlation in the NPS compared to full scan. In conclusion, noise
spatial structure is directionally dependent off-center and this may have an impact on object detectability in directional
In this work we apply the circle-and-line acquisition for the 256-detector row medical CT scanner. Reconstruction is based on the exact algorithm of the FBP type suggested recently by one of the co-authors. We derived equations for the cylindrical detector, common for medical CT scanners. To minimize hardware development efforts we use ramp-based reconstruction of the circle data. The line data provides an additional term that corrects the cone beam artifacts that are caused by the incompleteness of the circular trajectory. We illustrate feasibility of our approach using simulated data and real scanned data of the anthropomorphic phantom and evaluate stability of reconstruction to motion and misalignments during the scan. The additional patient dose from the line scan is relatively low compared to the circle scan. The proposed algorithm allows cone beam artifact-free reconstruction with large cone angle.
Proc. SPIE. 6316, Image Reconstruction from Incomplete Data IV
KEYWORDS: Fluctuations and noise, Sensors, Signal attenuation, Digital filtering, Medical research, Image filtering, Reconstruction algorithms, Binary data, Expectation maximization algorithms, Data analysis
Sinogram truncation is a common problem in tomographic reconstruction. Authors expand their previously published method of sinogram extension using decomposition into sinogram curves by using the adaptive convex filter. The main idea is to estimate the truncated parts of the projections of some object or patient using measured projections at different projection angles. This technique provides good estimation of the missing data near the edge of truncation. However, it is hard to estimate the outer edges of the truncated sinogram; in other words, the outer edge of the sinogram, and, consequently, reconstructed object, is invisible. To overcome this problem we introduce the adaptive convex filter that rounds off the outer portions of the extended sinogram, which tend to have a form of peak directed outwards. Here we assume that the truncated part of the reconstructed object has a round or elliptic shape, which holds with most clinical applications. The method automatically adjusts to the size of the truncated object, whether it is an arm or a part of torso.
Sinogram truncation is a common problem in tomographic reconstruction; it occurs when a scanned object or patient extends outside the scan field-of-view. The truncation artifact propagates from the edge of truncation towards the center, resulting in degraded image quality. Several methods have been proposed recently to reconstruct the image artifact-free within the scan FOV; however it is often necessary to recover image outside the scan FOV. We propose a novel truncation correction algorithm that accurately completes unmeasured data outside of the scan field-of-view, which allows us to extend the reconstruction field-of-view. Contrary to 1D extrapolation, we perform interpolation along the so-called sinogram curves. First, we propose an approach to parameterize the family of sinogram curves for efficient sinogram decomposition. Secondly, we propose two ways to estimate the truncated data outside the field-of-view. Both methods are combined for more accurate sinogram completion. Our evaluation shows the validity of our approach. Even objects completely outside the FOV can be accurately reconstructed using the proposed method. The proposed method can be used with any modality where sinogram truncation occurs, such as CT, C-arm, PET/CT, and SPECT.
SC939: Exact Cone Beam Reconstruction: Theory and Practice
This course provides attendees with basic working knowledge of the fundamentals of exact image reconstruction in cone beam CT. The course starts with the general theory, then we discuss various approaches to obtaining inversion formulae, and then we consider specific trajectories, such as helical and circle plus a curve. We include a discussion of implementation techniques, analysis of detector requirements and data usage. We will also discuss image quality of exact Katsevich-type (shift-invariant filtered-backprojection structure) reconstruction.
• Foundations of three-dimensional image reconstruction theory in computed tomography - Radon transform, cone beam transform, Grangeat's formula
• General reconstruction scheme - intersections of the source trajectory with Radon planes, weight function n, inversion of the cone beam transform
• Approaches to obtaining reconstruction formulae, including the Zou-Pan approach - Reconstruction on chords; Gelfand-Graev formula; Pack-Noo approach - Reconstruction on M-lines; and other approaches
• Trajectory-specific choice of the weight function for optimal reconstruction performance, both helical (1-PI, 3-PI, and Fractional-PI) and generalized circle-plus trajectories (open circle + line, and closed circle + curve)
• Implementation details including filtering lines rebinning and detector requirements
• Image quality