Monte Carlo Simulations of the GE Discovery Alcyone CZT SPECT Systems
- Medicinsk strålningsfysik, Lund
Publikation/Tidskrift/Serie: IEEE Transactions on Nuclear Science
Dokumenttyp: Artikel i tidskrift
Förlag: IEEE--Institute of Electrical and Electronics Engineers Inc.
Compact SPECT systems with cadmium zinc telluride (CZT) solid-state detectors with improved energy resolution and shorter acquisition times have recently been introduced. These systems have, however, different energy characteristics compared to NaI(Tl) crystal-based cameras; therefore, a need exists to develop new simulation models for these cameras. We modeled the charge transport within the CZT detectors for a GE Discovery 530c/570c SPECT system with multiple pinhole collimators employing the SIMIND Monte Carlo program and validated simulations against measurements. The incomplete charge collection between the anode and cathode in the pixilated CZT was modeled with the Hecht equation. The simulation also included charge-sharing effects across pixels due to physical interactions and charge diffusion. To validate our CZT-model, Tc-99m and I-123 point sources and a Tl-201 line source were acquired and measured. Measured energy spectra were compared with simulated energy spectra. The Monte Carlo simulated energy spectra agreed well with the experimental measurements within the photopeak, overestimated the k-edge x-ray escape peaks of cadmium and telluride, and slightly underestimated the remainder of the tail. Comparisons of system sensitivity and spatial resolution were also conducted for an array of point source locations with results showing excellent agreement. Lastly, to demonstrate a clinically realistic case, a simulation of an anthropomorphic phantom with a cardiac insert and an inferior defect was performed. Simulated projections were processed using the GE Xeleris software confirming the accuracy of the SIMIND geometry. We conclude that it is feasible to simulate the GE Discovery 530c/570c SPECT system using the SIMIND code.
- Radiology, Nuclear Medicine and Medical Imaging
- Monte Carlo
- ISSN: 0018-9499
E-post: michael [dot] ljungberg [at] med [dot] lu [dot] se