Search

Ordered subsets expectation maximization (OS-EM) is widely used to accelerate image reconstruction in single photon emission computed tomography (SPECT). Speedup of OS-EM over maximum likelihood expectation maximization (ML-EM) is close to the number of subsets used. Although a high number of subsets can shorten reconstruction times significantly, it can also cause severe image artifacts such as improper erasure of reconstructed activity if projections contain few counts. We recently showed that such artifacts can be prevented by using a count-regulated OS-EM (CR-OS-EM) algorithm which automatically adapts the number of subsets for each voxel based on the estimated number of counts that the voxel contributed to the projections. While CR-OS-EM reached high speed-up over ML-EM in high-activity regions of images, speed in low-activity regions could still be very slow. In this work we propose similarity-regulated OS-EM (SR-OS-EM) as a much faster alternative to CR-OS-EM. SR-OS-EM also automatically and locally adapts the number of subsets, but it uses a different criterion for subset regulation: the number of subsets that is used for updating an individual voxel depends on how similar the reconstruction algorithm would update the estimated activity in that voxel with different subsets. Reconstructions of an image quality phantom and in vivo scans show that SR-OS-EM retains all of the favorable properties of CR-OS-EM, while reconstruction speed can be up to an order of magnitude higher in low-activity regions. Moreover our results suggest that SR-OS-EM can be operated with identical reconstruction parameters (including the number of iterations) for a wide range of count levels, which can be an additional advantage from a user perspective since users would only have to post-filter an image to present it at an appropriate noise level.

Statistical reconstruction methods offer possibilities to improve image quality as compared with analytical methods, but current reconstruction times prohibit routine application in clinical and micro-CT. In particular, for cone-beam x-ray CT, the use of graphics hardware has been proposed to accelerate the forward and back-projection operations, in order to reduce reconstruction times. In the past, wide application of this texture hardware mapping approach was hampered owing to limited intrinsic accuracy. Recently, however, floating point precision has become available in the latest generation commodity graphics cards. In this paper, we utilize this feature to construct a graphics hardware accelerated version of the ordered subset convex reconstruction algorithm. The aims of this paper are (i) to study the impact of using graphics hardware acceleration for statistical reconstruction on the reconstructed image accuracy and (ii) to measure the speed increase one can obtain by using graphics hardware acceleration. We compare the unaccelerated algorithm with the graphics hardware accelerated version, and for the latter we consider two different interpolation techniques. A simulation study of a micro-CT scanner with a mathematical phantom shows that at almost preserved reconstructed image accuracy, speed-ups of a factor 40 to 222 can be achieved, compared with the unaccelerated algorithm, and depending on the phantom and detector sizes. Reconstruction from physical phantom data reconfirms the usability of the accelerated algorithm for practical cases.

Statistical reconstruction methods offer possibilities of improving image quality as compared to analytical methods, but current reconstruction times prohibit routine clinical applications. To reduce reconstruction times we have parallelized a statistical reconstruction algorithm for cone-beam x-ray CT, the ordered subset convex algorithm (OSC), and evaluated it on a shared memory computer. Two different parallelization strategies were developed: one that employs parallelism by computing the work for all projections within a subset in parallel, and one that divides the total volume into parts and processes the work for each sub-volume in parallel. Both methods are used to reconstruct a three-dimensional mathematical phantom on two different grid densities. The reconstructed images are binary identical to the result of the serial (non-parallelized) algorithm. The speed-up factor equals approximately 30 when using 32 to 40 processors, and scales almost linearly with the number of cpus for both methods. The huge reduction in computation time allows us to apply statistical reconstruction to clinically relevant studies for the first time.

Statistical methods for image reconstruction such as maximum likelihood expectation maximization (ML-EM) are more robust and flexible than analytical inversion methods and allow for accurate modelling of the photon transport and noise. Statistical reconstruction is prohibitively slow when applied to clinical x-ray cone-beam CT due to the large data sets and the high number of iterations required for reconstructing high resolution images. One way to reduce the reconstruction time is to use ordered subsets of projections during the iterations, which has been successfully applied to fan-beam x-ray CT. In this paper, we quantitatively analyse the use of ordered subsets in concert with the convex algorithm for cone-beam x-ray CT reconstruction, for the case of circular acquisition orbits. We focus on the reconstructed image accuracy of a 3D head phantom. Acceleration factors larger than 300 were obtained with errors smaller than 1%, with the preservation of signal-to-noise ratio. Pushing the acceleration factor towards 600 by using an increasing number of subsets increases the reconstruction error up to 5% and significantly increases noise. The results indicate that the use of ordered subsets can be extremely useful for cone-beam x-ray CT.

We describe a newly developed, accelerated Monte Carlo simulator of a small animal micro-CT scanner. Transmission measurements using aluminium slabs are employed to estimate the spectrum of the x-ray source. The simulator incorporating this spectrum is validated with micro-CT scans of physical water phantoms of various diameters, some containing stainless steel and Teflon rods. Good agreement is found between simulated and real data: normalized error of simulated projections, as compared to the real ones, is typically smaller than 0.05. Also the reconstructions obtained from simulated and real data are found to be similar. Thereafter, effects of scatter are studied using a voxelized software phantom representing a rat body. It is shown that the scatter fraction can reach tens of per cents in specific areas of the body and therefore scatter can significantly affect quantitative accuracy in small animal CT imaging.

Monte Carlo (MC) methods can accurately simulate scatter in X-ray imaging. However, when low noise scatter projections have to be simulated these MC simulations tend to be very time consuming. Rapid computation of scatter estimates is essential for several applications. The aim of the work presented in this paper is to speed up the estimation of noise-free scatter projections while maintaining their accuracy. Since X-ray scatter projections are often rather smooth, an approach is chosen whereby a short MC simulation is combined with a data fitting program that is robust to projection truncation and noise. This method allows us to estimate the smooth scatter projection rapidly. The speed-up and accuracy achieved by using the fitting algorithm were validated for the projection simulation of a small animal X-ray CT system. The acceleration that can be obtained over standard MC simulations is typically two orders of magnitude, depending on the accuracy required. The proposed approach may be useful for rapid simulation of patient and animal studies and for correction of the image-degrading effects of scatter in tomography.

Iterative maximum likelihood (ML) transmission computed tomography algorithms have distinct advantages over Fourier-based reconstruction, but unfortunately require increased computation time. The convex algorithm [1] is a relatively fast iterative ML algorithm but it is nevertheless too slow for many applications. Therefore, an acceleration of this algorithm by using ordered subsets of projections is proposed [ordered subsets convex algorithm (OSC)]. OSC applies the convex algorithm sequentially to subsets of projections. OSC was compared with the convex algorithm using simulated and physical thorax phantom data. Reconstructions were performed for OSC using eight and 16 subsets (eight and four projections/subset, respectively). Global errors, image noise, contrast recovery, and likelihood increase were calculated. Results show that OSC is faster than the convex algorithm, the amount of acceleration being approximately proportional to the number of subsets in OSC, and it causes only a slight increase of noise and global errors in the reconstructions. Images and image profiles of the reconstructions were in good agreement. In conclusion, OSC and the convex algorithm result in similar image quality but OSC is more than an order of magnitude faster.

Although the potential benefits of maximum likelihood reconstruction have been recognised for many years, the technique has only recently found widespread popularity in clinical practice. Factors which have contributed to the wider acceptance include improved models for the emission process, better understanding of the properties of the algorithm and, not least, the practicality of application with the development of acceleration schemes and the improved speed of computers. The objective in this article is to present a framework for applying maximum likelihood reconstruction for a wide range of clinically based problems. The article draws particularly on the experience of the three authors in applying an acceleration scheme involving use of ordered subsets to a range of applications. The potential advantages of statistical reconstruction techniques include: (a) the ability to better model the emission and detection process, in order to make the reconstruction converge to a quantitative image, (b) the inclusion of a statistical noise model which results in better noise characteristics, and (c) the possibility to incorporate prior knowledge about the distribution being imaged. The great flexibility in adapting the reconstruction for a specific model results in these techniques having wide applicability to problems in clinical nuclear medicine.

Abstract: This study aims at quantifying the 3-D variability in lithosphere strength of the south-eastern Canadian Cordillera and adjacent craton to the east. Strength is calculated in a forward manner, starting from rheological laws of brittle and ductile deformation. The work flow calculates a temperature model based on a multi-layer compositional model and subsequently estimates the strength distribution from both the compositional and temperature models. The temperature modeling involves numerical inversion and cubic spline algorithms which enable to impose boundary conditions that account better for strong lateral variations in lithosphere thickness and lateral heat flow. This addresses the lithosphere structure between the Canadian Cordillera and craton. The Canadian Cordillera is marked by a hot and thin lithosphere of ~60km thickness that stands in contrast to a cold and thick craton of ~170km thickness to the east, which consequently results in pronounced changes in bulk lithosphere strength. The high surface heat flow of the Cordillera interior and its contrast with the Foreland Basin can be reproduced by temperature models that combine elevated mantle heat flow due to the thin lithosphere and higher crustal heat production from magmatic intrusions. A series of rheological models, which examines the role of different temperature input models, composition and strain rate, shows that the first-order strength pattern is persistent. For the hot Canadian Cordillera, strength resides for >80% in the upper crust and with integrated lithosphere strength of 2.0–3.5⋅106 MPa⋅m. For the cold craton, the upper mantle provides >75% of the integrated strength with values of 4.0–8.0⋅106 MPa⋅m for the crust and 20–65⋅106 MPa⋅m for the lithosphere. Effective elastic thickness is estimated between 5 and 15km for the Cordillera and 40–80km for the craton. This illustrates the Cordillera to craton transition as a prominent rheological feature for upper mantle flow dynamics and spatial seismicity trends. [Copyright &y& Elsevier]