A collapsed-cone based transit EPID dosimetry method.

PURPOSE: To develop a transit-dose portal dosimetry method based on a commercial collapsed-cone algorithm. METHODS: A Varian Clinac21EX (Varian Medical Systems, USA), equipped with an amorphous-silicon EPID aS1000, was used. Dose calculations were performed with the collapsed-cone algorithm of Pinnacle3 v8.0 m (Philips Medical Systems, USA). A model for the energy of 6 MV was made in Pinnacle3 and afterwards validated for clinical use. A virtual phantom with different densities was contoured and superimposed on the patient images, simulating the presence of the EPID during the treatment. Corrections for photon spectral variations were introduced using Matlab (Mathworks, USA). Transit dosimetry was verified with an anthropomorphic phantom, on which different treatment fields were simulated in locations of skull, thorax and pelvis. In addition, a prostate treatment with IMRT was administered thereon. Dose distributions were compared with gamma index. RESULTS: The dose differences at the central point did not exceed 2%, except for the 20 x 20 cm2 field size centered in the skull. The model presented in this work, assumes that the dimensions of the solid water phantom, are infinite, except for the thickness. The mean values for the gamma index pass rates were 85.62% for (3%, 3 mm), 91.73% (4%, 3 mm) and 95.68% (5%, 3 mm). CONCLUSIONS: The value of 95% for γ (5%, 3 mm) can be established as the value below which the origin of the discrepancies should be investigated. It should be considered that the proposed method is complementary and not a substitute for pre-treatment dosimetry.