Application of Monte Carlo method for the calibration of an in vivo measurement system

Full text. Although great efforts have been made to constantly improve the physical phantoms to calibrate in vivo measurement systems, these phantoms represent a single average counting geometry and usually contain a uniform distribution of the radionuclide over the tissue substitute. The calibration method presented here employs a mathematical phantom, produced in the form of volume elements (voxels), obtained through Magnetic Resonance Images (MRI) of the human body. The calibration method uses Monte Carlo techniques to simulate the tissue contamination, to transport the photons through the tissues and to simulate the detection of the radiation. The program transports and detects photons between 35 and 2000 keV, and uses a format of 200 'slices' each of 277 x 148 picture elements (pixels) for the body representation. The angular and energy dependence of the mathematical detectors were determined using point sources of 137 Cs and 241 Am. a 3952 Bq source of 137 Cs was counted at nine positions with respect to the 8''x 4'' Nal(Tl) detector and a 29650 Bq source of 241 Am was counted at 3 positions with respect to the Phoswich detector. The Phoswich detector is shielded from low energy photons on its sides. The MC-VB program was run for 30000 photon arrivals at the detector. The results of the point source measurements for 137 Cs made with the 8''x 4'' Nal (Tl) detector allow the calculation of the simulated calibration factor, which is defined here as the net counts/minute calculated from the Monte Carlo (MC-CPM) divided by the measured net count / minute (measured cpm). The simulated calibration factor is 1.17 +/- 0.03 for 137 Cs using the 8''x 4'' Nal(Tl) detector and 1.38 +/- 0.04 for 241 Am using the Phoswich detector