ABSTRACT
Objective To study the accuracy of collapsed cone convolution ( CCC) and anisotropic analytical algorithm ( AAA) in dosimetric calculation on the air cavity interface. Methods A BEAMnrc/EGSnrc Monte Carlo ( MC ) simulation was performed on a Varian Trilogy linear accelerator. The IBA Dosimetry “blue phantom” 3D scanning system was used to verify the accuracy and reliability of the MC simulation. Central axis depth dose distribution and lateral dose profile in a water?equivalent phantom with variously sized air cavities were calculated by CCC and AAA. The obtained depth dose distribution and lateral dose profile were compared with those by MC simulation and EBT2 film, respectively. Results Both CCC and AAA overestimated the dose on the air cavity interface. In spite of some errors, CCC had a higher accuracy than AAA. The errors were mainly related to computational grid, field size, photon energy, cavity size, and the number of fields. Conclusion Electronic disequilibrium on the air cavity interface should be taken into account when CCC and AAA are used for dosimetric calculation in treatment planning system.
ABSTRACT
Cyberknife with small field size is more difficult and complex for dosimetry compared with conventional radiotherapy due to electronic disequilibrium, steep dose gradients and spectrum change of photons and electrons. The purpose of this study demonstrate the usefulness of Geant4 as verification tool of measurement dose for delivering accurate dose by comparing measurement data using the diode detector with results by Geant4 simulation. The development of Monte Carlo Model for Cyberknife was done through the two-step process. In the first step, the treatment head was simulated and Bremsstrahlung spectrum was calculated. Secondly, percent depth dose (PDD) was calculated for six cones with different size, i.e., 5 mm, 10 mm, 20 mm, 30 mm, 50 mm and 60 mm in the model of water phantom. The relative output factor was calculated about 12 fields from 5 mm to 60 mm and then it compared with measurement data by the diode detector. The beam profiles and depth profiles were calculated about different six cones and about each depth of 1.5 cm, 10 cm and 20 cm, respectively. The results about PDD were shown the error the less than 2% which means acceptable in clinical setting. For comparison of relative output factors, the difference was less than 3% in the cones lager than 7.5 mm. However, there was the difference of 6.91% in the 5 mm cone. Although beam profiles were shown the difference less than 2% in the cones larger than 20 mm, there was the error less than 3.5% in the cones smaller than 20 mm. From results, we could demonstrate the usefulness of Geant4 as dose verification tool.