Two approaches to implementing projector–backprojector pairs for 3D reconstruction from Compton scattered data

Kim, Soo Mee; Lee, Jae Sung; Lee, Mi No; Lee, Ju Hahn; Lee, Chun Sik; Kim, Chan Hyeong; Lee, Dong Hoon; Lee, Soo Jin

doi:10.1016/j.nima.2006.10.076

Cited by 27 publications

(17 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other detector factors such as energy resolution and spatial resolution are not considered in this study and an otherwise 'perfect' camera was assumed. Exact interaction positions and energy depositions in the two detectors for successful photon events were saved to output data files and images of the point source were reconstructed from the simulated data using a simple back projection algorithm [15].…”

Section: Methodsmentioning

confidence: 99%

GEANT4 simulation of the effects of Doppler energy broadening in Compton imaging

Uche

Cree

Round

2011

Australas Phys Eng Sci Med

View full text Add to dashboard Cite

A Monte Carlo approach was used to study the effects of Doppler energy broadening on Compton camera performance. The GEANT4 simulation toolkit was used to model the radiation transport and interactions with matter in a simulated Compton camera. The low energy electromagnetic physics model of GEANT4 incorporating Doppler broadening developed by Longo et al. was used in the simulations. The camera had a 9 × 9 cm scatterer and a 10 × 10 cm absorber with a scatterer to-absorber separation of 5 cm. Modelling was done such that only the effects of Doppler broadening were taken into consideration and effects of scatterer and absorber thickness and pixelation were not taken into account, thus a 'perfect' Compton camera was assumed. Scatterer materials were either silicon or germanium and the absorber material was cadmium zinc telluride. Simulations were done for point sources 10 cm in front of the scatterer. The results of the simulations validated the use of the low energy model of GEANT4. As expected, Doppler broadening was found to degrade the Compton camera imaging resolution. For a 140.5 keV source the resulting full-width-at-half-maximum (FWHM) of the point source image without accounting for Doppler broadening and using a silicon scatterer was 0.58 mm. This degraded to 7.1 mm when Doppler broadening was introduced and degraded further to 12.3 mm when a germanium scatterer was used instead of silicon. But for a 511 keV source, the FWHM was better than for a 140 keV source. The FWHM improved to 2.4 mm for a silicon scatterer and 4.6 mm for a germanium scatterer. Our result for silicon at 140.5 keV is in very good agreement with that published by An et al.

show abstract

Section: Methodsmentioning

confidence: 99%

GEANT4 simulation of the effects of Doppler energy broadening in Compton imaging

Uche

Cree

Round

2011

Australas Phys Eng Sci Med

View full text Add to dashboard Cite

show abstract

“…Note that for these simulations the scatterer was still assumed perfect. The simulated energy and position data were then processed on an event basis using a listmode back-projection algorithm whose ellipse equation was derived using the steps provided by [21]. The algorithm analyzed each parameter of the event data to reconstruct a source image.…”

Section: Resolution Issuesmentioning

confidence: 99%

A Monte Carlo evaluation of three Compton camera absorbers

Uche

Round

Cree

2011

Australas Phys Eng Sci Med

View full text Add to dashboard Cite

We present a quantitative study on the performance of cadmium zinc telluride (CZT), thallium-doped sodium iodide (NaI(Tl)) and germanium (Ge) detectors as potential Compton camera absorbers. The GEANT4 toolkit was used to model the performance of these materials over the nuclear medicine energy range. CZT and Ge demonstrate the highest and lowest efficiencies respectively. Although the best spatial resolution was attained for Ge, its lowest ratio of single photoelectric to multiple interactions suggests that it is most prone to inter-pixel cross-talk. In contrast, CZT, which demonstrates the least positioning error due to multiple interactions, has a comparable spatial resolution with Ge. Therefore, we modelled a Compton camera system based on silicon (Si) and CZT as the scatterer and absorber respectively. The effects of the detector parameters of our proposed system on image resolution were evaluated and our results show good agreement with previous studies. Interestingly, spatial resolution which accounted for the least image degradation at 140.5 keV became the dominant degrading factor at 511 keV, indicating that the absorber parameters play some key roles at higher energies. The results of this study have validated the predictions by An et al. which state that the use of a higher energy gamma source together with reduction of the absorber segmentation to sub-millimetre could achieve the image resolution of 5 mm required in medical imaging.

show abstract

“…(4) and (5). In both equations, the geometric system matrix, H ij , is implemented by ray-tracing method described in our previous work [13]. In Eq.…”

Section: Lmosem Reconstruction Algorithm Including Gaussian Resolutiomentioning

confidence: 99%

Study on listmode OSEM reconstruction including image-space resolution recovery techniques for Compton camera

Kim

Lee

Park

et al. 2010

2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro

View full text Add to dashboard Cite

Although Compton camera may has a great potential as next generation imaging modality comparing to SPECT and PET, its fully three-dimensional image reconstruction requires the considerable computational burden and the spatial resolution is suffered from the various physical phenomena arising during detection process. In this study, we investigated the accelerated statistical image reconstruction in which system matrix included a resolution recovery (RR) technique. We considered 3D Gaussian resolution model for the integrated angular and geometric uncertainties. The angular uncertainty is closely related to the limited energy resolution of the Compton camera and Doppler broadening and the geometric uncertainty is due to the segmented detectors. For RR, the 3D Gaussian resolution model is incorporated into listmode OSEM (LMOSEM) using image-space convolution operation. We investigated two different RR approaches: one (denoted by LMOSEM-RR F ) is when the convolution is only performed in forward projection step, and the other (denoted by LMOSEM-RR FB ) is when it is performed in both forward and backward projection steps. The simulation results showed that both RR approaches gave an improvement on spatial resolution for the resolution-degraded data due to both angular and geometric uncertainties. Although LMOSEM-RR F provided better resolution than LMOSEM-RR FB , LMOSEM-RR FB could still useful for low counting statistics in measurement.

show abstract

Two approaches to implementing projector–backprojector pairs for 3D reconstruction from Compton scattered data

Cited by 27 publications

References 7 publications

GEANT4 simulation of the effects of Doppler energy broadening in Compton imaging

GEANT4 simulation of the effects of Doppler energy broadening in Compton imaging

A Monte Carlo evaluation of three Compton camera absorbers

Study on listmode OSEM reconstruction including image-space resolution recovery techniques for Compton camera

Contact Info

Product

Resources

About