Title :
Image reconstruction for PETbox, a benchtop preclinical PET tomograph
Author :
Bao, Qinan ; Rannou, Fernando R. ; Taschereau, Richard ; Stout, David B. ; Chatziioannou, Arion F.
Author_Institution :
David Geffen Sch. of Med., Crump Inst. for Mol. Imaging, UCLA, Los Angeles, CA, USA
fDate :
Oct. 24 2009-Nov. 1 2009
Abstract :
PETbox, an integrated low-cost benchtop preclinical PET scanner dedicated to high throughput quantitative mouse studies is currently under development at the Crump Institute for Molecular Imaging, UCLA. The system employs two opposing flat panel BGO detector blocks on a static gantry and produces a limited number of projection angles around a central antero-posterior (AP) view. Iterative reconstruction based on a Maximum Likelihood and Expectation Maximization (ML-EM) algorithm was developed with incorporation of a probability matrix, which takes into account the detection probability and inter-crystal scattering. Point sources at different positions were simulated with the GATE software package and showed a peak sensitivity of 3.87% at a 150-650 keV energy window. When applying a sensitivity correction map, the image values of the reconstructed point sources at different positions are within a 5.3% standard deviation, indicating good quantification accuracy. The spatial resolution of point sources is largely uniform along the coronal direction across the field of view (FOV) with worse resolution along the direction orthogonal to the detector blocks, which is mainly due to the reduced spatial sampling in that direction. A voxelized digital mouse (Moby phantom) was also simulated and then iteratively reconstructed. Attenuation and non-uniform sensitivity corrections were applied. Quantification accuracy was evaluated with the reconstructed Moby data and two analytically simulated data sets. To a first approximation, reasonably good quantification accuracy was obtained for most important organs with the PETbox reconstructed image. This evaluation indicates that the integrated detector system design, reconstruction algorithm and attenuation correction work well, which is very important for quantitative pharmacokinetic and pharmacodynamic studies. In-vivo mouse studies were also acquired and reconstructed. The obtained images proved the capability of PETbox as a dedicat- - ed mouse scanner for both dynamic and static acquisitions.
Keywords :
image reconstruction; phantoms; positron emission tomography; solid scintillation detectors; BGO detector; GATE software package; Moby phantom; PETbox; benchtop preclinical PET tomograph; central antero-posterior view; expectation maximization algorithm; image reconstruction; in-vivo mouse studies; inter-crystal scattering; iterative reconstruction; maximum likelihood algorithm; mouse scanner; pharmacodynamic studies; point sources; probability matrix; quantification accuracy; quantitative pharmacokinetic studies; spatial resolution; standard deviation; voxelized digital mouse; Attenuation; Detectors; Image reconstruction; Iterative algorithms; Maximum likelihood detection; Mice; Molecular imaging; Positron emission tomography; Spatial resolution; Throughput;
Conference_Titel :
Nuclear Science Symposium Conference Record (NSS/MIC), 2009 IEEE
Conference_Location :
Orlando, FL
Print_ISBN :
978-1-4244-3961-4
Electronic_ISBN :
1095-7863
DOI :
10.1109/NSSMIC.2009.5401967