Premium
SU‐GG‐I‐144: Validation of a Monte Carlo Model of the PET Component of the Gemini GXL PET/CT
Author(s) -
Lavoie G,
Martel C,
Després P,
Lacroix F,
Chen Y,
Carrier J
Publication year - 2008
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.2961542
Subject(s) - monte carlo method , imaging phantom , physics , extrapolation , radius , offset (computer science) , image resolution , nuclear medicine , sensitivity (control systems) , transverse plane , optics , computational physics , mathematics , statistics , electronic engineering , computer science , medicine , computer security , structural engineering , programming language , engineering
Purpose: To validate a Monte Carlo model of the PET component of the Gemini GXL PET/CT scanner (Philips Medical Systems) developed using the GATE simulation toolkit. Method and Materials: The precise PET geometry was reproduced with the GATE toolkit, including the 17 864 GSO(Zr) crystals with dimensions of 4 mm × 6 mm × 30 mm. A series of standard tests from the NEMA NU 2‐2001 protocol were simulated with the Monte Carlo model and compared to experimental data: sensitivity, spatial resolution, scatter fraction, and noise equivalent count rate. The sensitivity is evaluated through an extrapolation technique both in the middle and at a 10 cm radial offset of a uniform line source surrounded by known absorbers. The spatial resolution is given by the full width at half‐maximum amplitude of the point spread function in the transverse and axial directions for compact radioactive sources in a glass capillary. The scatter fraction and noise equivalent count (NEC) rate are evaluated in a cylindrical phantom with variable radioactivity in a line source. Results: For the sensitivity test, the simulations for both central and offset positions overestimate the measured sensitivity values by 4%. For the resolution test, the difference between experimental and simulated data is between 0.4 mm (radial resolution at 10 cm radius) and 1.3 mm (transverse resolution at 1 cm radius). For the scatter fraction, the simulated results are within 0.5% of the experimental data. For the count rate tests, the peak NEC rate is at (62.8+/−0.5) kcps in the simulation, which is 4% lower than the measurements. Globally, the agreement for the count rate is always better than 5% for all activities less than 17 kBq/cc. Conclusion: The PET Monte Carlo model reproduces experimental validation data within acceptable limits. It is suitable for PET testing without experimental use of FDG.