Scatter Correction in PET Using the Transport Equation

Kösters, Thomas; Natterer, Frank; Wübbeling, Frank

doi:10.1109/nssmic.2006.353713

Cited by 5 publications

(4 citation statements)

References 6 publications

(6 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, we estimate that the so-called EM-TV algorithm [3] applied to 3D PET and an advanced scatter correction method [5] will both be one to two orders of magnitude more compute-intensive than the current algorithm. In order to use such algorithms, clusters of multi-core processors will probably be the architecture to target.…”

Section: Resultsmentioning

confidence: 99%

Parallel medical image reconstruction: from graphics processing units (GPU) to Grids

et al. 2010

View full text Add to dashboard Cite

We present and compare a variety of parallelization approaches for a realworld case study on modern parallel and distributed computer architectures. Our case study is a production-quality, time-intensive algorithm for medical image reconstruction used in computer tomography (PET). We parallelize this algorithm for the main kinds of contemporary parallel architectures: shared-memory multiprocessors, distributed-memory clusters, graphics processing units (GPU) using the CUDA framework, the Cell processor and, finally, how various architectures can be accessed in a distributed Grid environment. The main contribution of the paper, besides the parallelization approaches, is their systematic comparison regarding four important criteria: performance, programming comfort, accessibility, and cost-effectiveness. We report results of experiments on particular parallel machines of different architectures that confirm the findings of our systematic comparison.

show abstract

Section: Resultsmentioning

confidence: 99%

Parallel medical image reconstruction: from graphics processing units (GPU) to Grids

et al. 2010

View full text Add to dashboard Cite

show abstract

“…Some additional positive considerations in this direction follow from the positron emission tomography model suggested in [14]. The model is an integral transform whose kernel is the product of the solutions of the radiation transfer equation.…”

Section: Numerical Experiments With the Inhomogeneity Indicatormentioning

confidence: 99%

“…In recent years, there has been developed an approach directed to construction of novel mathematical models of PET which allow us using in a certain way the information containing in the scattered signal, instead of its filtering [12][13][14].…”

mentioning

confidence: 99%

Numerical experiments with the inhomogeneity indicator in positron emission tomography

Yarovenko

2012

J. Appl. Ind. Math.

View full text Add to dashboard Cite

By means of numerical methods, the question is studied of applicability of the inhomogeneity indicator in positron emission tomography. The signal registered by the tomograph is described in terms of an imitation model using the Monte Carlo method. The possibility is demonstrated of the effective use of the inhomogeneity indicator for solving the problem under consideration. Some numerical results are presented in graphical form for reconstructing the boundaries of unknown activity sources.The foundation of positron emission tomography (PET) are based on the idea of registration of two oppositely directed gamma-quanta appearing as a result of annihilation of the positrons emitted by a special agent introduced into the patient's body. Under this approach, the functioning of the tomograph's detectors function, depending on one another. If two detectors simultaneously register a signal (i.e., the so-called coincidence happens) then it is assumed that the annihilation point is situated on the line joining the detectors. This line is called the line of response. The registration of a pair of photons allows us to proceed without the collimators in determining the direction of the gamma-quanta movement and is a specific filter which helps distinguishing exactly those trajectories of photons that carry the necessary information on the distribution of the activity source in the matter [1].The majority of scanners can function both in the slice-by-slice mode (when the axial collimation is created with the help of the special tungsten plates called the septa), and in the three-dimensional mode (when the septa are pulled in and the coincidences are registered between all possible pairs of detectors). The usage of collimation leads to diminishing the random components in the detected signal. At the same time, a three-dimensional PET system, allowing registering the coincidences between any rings of detectors, leads to an eight-fold increase of the number of registered coincidences in comparison with the slice-by-slice regime [2], which amplifies the sensitivity of the tomograph and allows faster accumulating the necessary volume of information. In connection with this, the considered research area is most promising [1]. However, increase of the total number of registered events leads inevitably to increase of the quantity of random events whose effect on the reconstruction quality is negative. For example, registration of the scattered photons results in distinguishing some false response lines. This circumstance, in turn, leads to the necessity of development of new information processing methods [1].The main body of research in this area is directed to the construction of various empirical formulas and methods to allow estimating the contribution of the scattered radiation into the signal registered by the detectors and distinguishing the nonscattered (ballistic) portion of radiation [3][4][5][6][7][8][9][10][11]. After that the tomography problem is reduced to a simple inversion of the Radon transform, and the theory in th...

show abstract

“…In this paper, we study Positron Emission Tomography (PET ) reconstruction, where one of the most popular, but also most time-consuming algorithms-the list-mode OSEM algorithm-requires several hours on a common PC in order to compute a 3D reconstruction. With advanced algorithms that incorporate more physical aspects of the PET process, computation times are rising even further [1]. This motivates the parallelization of the algorithm on multiprocessor machines [2].…”

Section: Introductionmentioning

confidence: 99%

Communication Optimization for Medical Image Reconstruction Algorithms

Hoefler

Schellmann

Gorlatch

et al. 2008

Recent Advances in Parallel Virtual Machine and Message Passing Interface

View full text Add to dashboard Cite

Abstract. This paper presents experiences and results obtained in optimizing the parallel communication performance of a production-quality medical image reconstruction application. The fundamental communication operations in the application's principal algorithm are collective reductions. The overhead of these operations was reduced by transforming the algorithm to overlap its computation and communication. Several different approaches to communication progress were studied, both user-directed and asynchronous. Experimental results comparing the new approach to the previous implementation show overall application performance improvements of up to 8%, when run on 32 nodes.

show abstract

Scatter Correction in PET Using the Transport Equation

Cited by 5 publications

References 6 publications

Parallel medical image reconstruction: from graphics processing units (GPU) to Grids

Parallel medical image reconstruction: from graphics processing units (GPU) to Grids

Numerical experiments with the inhomogeneity indicator in positron emission tomography

Communication Optimization for Medical Image Reconstruction Algorithms

Contact Info

Product

Resources

About