Fast and memory‐efficient Monte Carlo‐based image reconstruction for whole‐body PET

Long, Zhang; Staelens, Steven; Holen, Roel Van; Beenhouwer, Jan De; Verhaeghe, Jeroen; Kawrakow, I.; Vandenberghe, Stefaan

doi:10.1118/1.3455287

Cited by 38 publications

(40 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The resulting matrices have far fewer non-zero elements, and the approach provides the flexibility of being able to choose the most appropriate method (analytic, experimental or Monte Carlo) for obtaining each of the aforementioned components. Alternatively, by exploiting the geometric symmetry of the scanner (axial and planar symmetries) (Zhang et al 2010) as well as by using a polar rather than cubic voxelisation method (Cabello and Rafecas 2011), the size of the system matrix can be reduced. In addition, Harraiz et al have proposed the concept of quasi-symmetry (Herraiz et al 2006), in which a number of quasi-equivalent LoR classes are constructed to categorise LoRs associated with different rings along the axial direction for a 3D small animal PET scanner, providing further reductions in system matrix storage requirements.…”

Section: Introductionmentioning

confidence: 99%

A new virtual ring-based system matrix generator for iterative image reconstruction in high resolution small volume PET systems

Safavi-Naeini

Franklin

et al. 2015

Phys. Med. Biol.

View full text Add to dashboard Cite

A common approach to improving the spatial resolution of small animal PET scanners is to reduce the size of scintillation crystals and/or employ high resolution pixellated semiconductor detectors. The large number of detector elements results in the system matrix-an essential part of statistical iterative reconstruction algorithms-becoming impractically large. In this paper, we propose a methodology for system matrix modelling which utilises a virtual single-layer detector ring to greatly reduce the size of the system matrix without sacrificing precision. Two methods for populating the system matrix are compared; the first utilises a geometrically-derived system matrix based on Siddon's ray tracer method with the addition of an accurate detector response function, while the second uses Monte Carlo simulation to populate the system matrix. The effectiveness of both variations of the proposed technique is demonstrated via simulations of PETiPIX, an ultra high spatial resolution small animal PET scanner featuring high-resolution DoI capabilities, which has previously been simulated and characterised using classical image reconstruction methods. Compression factors of × 5 10 7 and × 2.5 10 7 are achieved using this methodology for the system matrices produced using the Institute of Physics and Engineering in MedicineContent from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. geometric and Monte Carlo-based approaches, respectively, requiring a total of 0.5-1.2 GB of memory-resident storage. Images reconstructed from Monte Carlo simulations of various point source and phantom models, produced using system matrices generated via both geometric and simulation methods, are used to evaluate the quality of the resulting system matrix in terms of achievable spatial resolution and the CRC, CoV and CW-SSIM index image quality metrics. The Monte Carlo-based system matrix is shown to provide the best image quality at the cost of substantial one-off computational effort and a lower (but still practical) compression factor. Finally, a straightforward extension of the virtual ring method to a three dimensional virtual cylinder is demonstrated using a 3D DoI PET scanner.

show abstract

Section: Introductionmentioning

confidence: 99%

A new virtual ring-based system matrix generator for iterative image reconstruction in high resolution small volume PET systems

Safavi-Naeini

Franklin

et al. 2015

Phys. Med. Biol.

View full text Add to dashboard Cite

show abstract

“…Theo retical studies, such as the one presented in [9], suggests a number of simulated events based on the number of measure ments taken. Other authors have used empirical approaches to determine the number of simulated events [10]- [12], and a combination of theoretical and empirical methods has been used in [7].…”

Section: Introductionmentioning

confidence: 99%

“…The SRM can be either measured experimentally [2], [3], calculated using analytical methods [4], [5] or Monte Carlo (MC) methods [6], [7], or obtained using combinations of these methods [8]. MC and computing time necessary to calculate the SRM.…”

Section: Introductionmentioning

confidence: 99%

System response matrix denoising in measurement space in positron emission tomography

Cabello

Rafecas

2011

2011 IEEE Nuclear Science Symposium Conference Record

View full text Add to dashboard Cite

The system response matrix (SRM) maps measure ment space with image space in emission tomography. The quality of the SRM is of significant importance in statistical iterative methods. Different approaches are available to com pute the SRM. Generically, high image quality and accurate quantitative information depends on the accuracy and precision of the SRM elements. There is growing interest on Monte Carlo (MC) methods to compute the SRM due to the high accuracy achieved. The main drawback is that the SRM elements precision is strongly related with the number of events simulated, which depends on the simulation time. Long MC simulations produce low noiseJhigh precision SRM elements. In this work we investigate the impact of a denoising filter applied to the SRM elements of a short-time simulated SRM on the image quality and quantitative information accuracy. Results show that a ID mean filter applied longitudinally to profiles extracted from every volume of response produces better noise performance than the unfiltered SRM with ten times higher number of simulated events, at the cost of slightly degraded resolution. These results are further compared with post-reconstruction filtered images obtained with the unaltered SRM, demonstrating a larger level of spatial resolution degradation compared to those obtained with the filtered SRM.Index Terms-Nuclear imaging, PET, statistical iterative recon struction, Monte Carlo, system response matrix, denoising filter.

show abstract

“…Monte Cario simulation has been widely used to estimate accurately the system response for data projection without the inherent complexity of analytical models [2][3][4]. For this purpose, validated PET simulations tools like GATE [5] or PeneloPET [6] have been used.…”

Section: Introduction Smentioning

confidence: 99%

Projector model for efficient list-mode reconstruction in PET scanners with parallel planar detectors

Ortuño

Sportelli

Guerra

et al. 2011

2011 IEEE Nuclear Science Symposium Conference Record

View full text Add to dashboard Cite

Abstrcwt-We have developed a new projector model speciflcally tailored for fast list-mode tomographic reconstructions in Positrón emission tomography (PET) scanners with parallel planar detectors. The model provides an accurate estimation of the probability distribution of coincidence events defined by pairs of scintillating crystals. This distribution is parameterized with 2D elliptical Gaussian functions defined in planes perpendicular to the main axis of the tube of response (TOR). The parameters of these Gaussian functions have been obtained by fitting Monte Cario simulations that include positrón range, acolinearity of gamma rays, as well as detector attenuation and scatter effects. The proposed model has been applied efficiently to list-mode reconstruction algorithms. Evaluation with Monte Cario simulations over a rotating high resolution PET scanner indicates that this model allows to obtain better recovery to noise ratio in OSEM (ordered-subsets, expectationmaximization) reconstruction, if compared to list-mode reconstruction with symmetric circular Gaussian TOR model, and histogram-based OSEM with precalculated system matrix using Monte Cario simulated models and symmetries.

show abstract

Fast and memory‐efficient Monte Carlo‐based image reconstruction for whole‐body PET

Abstract: The results indicated that the proposed framework could be used for this PET scanner with improved image quality. This method could also be applied to other state-of-the-art whole-body PET scanners and preclinical PET scanners with a similar shape.

Cited by 38 publications

References 46 publications

A new virtual ring-based system matrix generator for iterative image reconstruction in high resolution small volume PET systems

A new virtual ring-based system matrix generator for iterative image reconstruction in high resolution small volume PET systems

System response matrix denoising in measurement space in positron emission tomography

Projector model for efficient list-mode reconstruction in PET scanners with parallel planar detectors

Contact Info

Product

Resources

About