A1 Vertaisarvioitu alkuperäisartikkeli tieteellisessä lehdessä
Acceleration of Monte Carlo-based scatter compensation for cardiac SPECT
Tekijät: Sohlberg A, Watabe H, Iida H
Kustantaja: IOP PUBLISHING LTD
Julkaisuvuosi: 2008
Journal: Physics in Medicine and Biology
Tietokannassa oleva lehden nimi: PHYSICS IN MEDICINE AND BIOLOGY
Lehden akronyymi: PHYS MED BIOL
Vuosikerta: 53
Numero: 14
Aloitussivu: N277
Lopetussivu: N285
Sivujen määrä: 9
ISSN: 0031-9155
DOI: https://doi.org/10.1088/0031-9155/53/14/N02
Tiivistelmä
Single proton emission computed tomography (SPECT) images are degraded by photon scatter making scatter compensation essential for accurate reconstruction. Reconstruction-based scatter compensation with Monte Carlo (MC) modelling of scatter shows promise for accurate scatter correction, but it is normally hampered by long computation times. The aim of this work was to accelerate the MC-based scatter compensation using coarse grid and intermittent scatter modelling. The acceleration methods were compared to un-accelerated implementation using MC-simulated projection data of the mathematical cardiac torso (MCAT) phantom modelling (99m)Tc uptake and clinical myocardial perfusion studies. The results showed that when combined the acceleration methods reduced the reconstruction time for 10 ordered subset expectation maximization (OS-EM) iterations from 56 to 11 min without a significant reduction in image quality indicating that the coarse grid and intermittent scatter modelling are suitable for MC-based scatter compensation in cardiac SPECT.
Single proton emission computed tomography (SPECT) images are degraded by photon scatter making scatter compensation essential for accurate reconstruction. Reconstruction-based scatter compensation with Monte Carlo (MC) modelling of scatter shows promise for accurate scatter correction, but it is normally hampered by long computation times. The aim of this work was to accelerate the MC-based scatter compensation using coarse grid and intermittent scatter modelling. The acceleration methods were compared to un-accelerated implementation using MC-simulated projection data of the mathematical cardiac torso (MCAT) phantom modelling (99m)Tc uptake and clinical myocardial perfusion studies. The results showed that when combined the acceleration methods reduced the reconstruction time for 10 ordered subset expectation maximization (OS-EM) iterations from 56 to 11 min without a significant reduction in image quality indicating that the coarse grid and intermittent scatter modelling are suitable for MC-based scatter compensation in cardiac SPECT.
Turun yliopiston Tutkimustietojärjestelmän portaali