Room: AAPM ePoster Library
To determine the offset, gain and curvature of jaws and MLC used to correct the nominal field size (FS) for a beam model of an Elekta Agility accelerator in RayStation.
FS ranging from 1x1 to 24x24 were measured with portal imaging for three photon energies (6, 10, 18 MV) on four matched accelerators. Portal images were analyzed with an in-house program. The difference between the measured and nominal FS was calculated and plotted as a function of nominal FS for all energy-axis combinations (total of six curves). A 2nd order polynomial curve adjustment was performed on each curve to obtain the offset, gain and curvature. The statistical significance of the fitting parameters and the correlation were evaluated.
All presented values are shown as ([estimate]±[standard error]). Y-jaws gain and curvature were significant (p<0.05) for all energies. Y-jaws gains were (–0.0092±0.0028), (–0.0095±0.0029) and (–0.0096±0.0038), and Y-jaws curvatures were (0.00070±0.00022) [1/cm], (0.00071±0.00023) [1/cm] and (0.00080±0.00030) [1/cm] for 6, 10 and 18 MV, respectively. Y-jaws offset was significant only for 10 MV with a value of (0.021±0.007) cm. Gain and curvature values for MLC x-position were not significant for any energy. MLC x-position offset was significant for 6 MV and 10 MV with values of (0.045±0.018) cm and (0.046±0.018) cm, respectively. The correlation for all six curves was weak (R^2=0.61).
RayStation allows to correct the effective jaws/MLC positions to account for position-dependant correction of FS. For an Elekta Agility, these corrections were either non significant or weakly correlated due to high variability in FS measurements. Gain and curvature could be set to zero for all axis and energies with negligible impact on the model. Offsets, although non significant for FS correction, could be used to achieved other modeling goals, such as output adjustment for IMRT plan.