When calculating the irradiation field,
it was found that phantom thicknesses of 1 mm,
5 mm,
and 10 mm were necessary to achieve electron equilibration for 100-300 keV,
400-1000 keV,
and 1500-2000 keV photons,
respectively.
Figure 8 shows the results of the consistency check for D and Kcol. The vertical axis shows that D/Kcol,
so D/Kcol=1 is the ideal value. The corresponding values of “Air/Air” (blue) and “PMMA/PMMA” (green) deviate within the range of 1±0.05. On the other hand,
the corresponding value of “Air/PMMA” (red) is systematically lower than 1,
therefore,
they are included in the range between 0.9 and 1.
Left graph in Fig.
9 shows the calculation efficiency for the condition of “Air/PMMA” for 100-2000 keV photons. The calculation efficiency becomes 70-90% for 100 keV and approximately 10% for 2000 keV. For 100 keV photons,
the calculation efficiency for 1 mm and 5 mm thicknesses are about 20% larger than those of 10 mm and 15 mm because a phantom thickness of 1 mm is sufficient for 100 keV and additional thickness results in attenuation. When increasing photon energy up to 1000 keV,
the differences become small. We determined that proper phantom thickness should be applied based on photon energy.
The right graph in Fig.
9 shows the fraction of scattered rays for the condition of “Air/PMMA” for 100-2000 keV photons. For the 1mm thick phantom,
the fraction of scattered X-rays varies approximately from 1.1% to 1.6% for photon energies from 100 keV to 300 keV. The 5 mm thick phantom varies approximately 2.7% to 4.7% for photon energies from 100 keV to 800 keV. The 10 mm thick phantom varies approximately 3.1% to 6.6% for photon energies from 100 keV to 2000 keV. Phantom thickness of 15 mm varies approximately 4.3% to 7.3% for photon energies of 100 keV to 2000 keV. As clearly seen in the graph,
the thicker the phantom used,
the more intense the scattered rays were generated.
Efficiency of Monte-Carlo simulation becomes 70-90% for 100 keV and approximately 10% for 2000 keV. These values are much higher than those without phantoms. Fraction of scattered rays becomes several percentages.
Moreover we estimated the accuracy of the simulation. From these results,
we evaluated that the accuracy of our system is approximately 10%.
Finally,
we demonstrate the ability of the proposed method. Figure 10 shows results from an additional simulation,
in which three different irradiation systems are applied; system A is the proposed system,
in system B the detection region is partially covered with a phantom and sides are not covered,
and in system C the detection region is not covered and the phantom is just placed in front of the detection region. The second condition is similar to the previously proposed practical experimental irradiation system which is valuable to achieve secondary electron equilibration [19]. The graph shows the availability of the proposed system; namely,
D/Kcol value of the proposed system (system A: red) is approximately 1,
but for systems B (green) and C the systematical value is smaller. The phenomenon indicates the importance of covering the detection region completely.