Degree Name

Doctor of Philosophy (PhD)


School of Engineering Physics - Faculty of Engineering


In 2002 when this research started the brief of the project was to produce streamlined checks of planar dose maps delivered by IMRT fields to film. At this time no other centre in Australia had a protocol for checking accuracy of RTP planned RT dose distributions. While many US centers have been checking IMRT distributions, there is still no standard protocol for these checks. By the end of this project in 2005, 13 IMRT patient treatments had been successfully checked and this centre remains the only centre to have treated IMRT patients in Australia using the pinnacle RTP planning computer platform. Early film dose maps revealed dose spikes due to MLC matchline effects. These matchlines were due to Varian MLC leaf ends sometimes matching other segment neighbors and were not predicted using pinnacle RTP until version 7.4 available about 2 months prior to the end of this project cycle. Verifying a radiation treatment planning (RTP) computer's IMRT calculation was the first task for this thesis. Planar dose maps (dose in water perpendicular to the beam [cGy/MU]) were compared with beam dose distributions measured using films (XV and EDR) at various depths. The RTP computer and film measurements agreed within &#;3% within the inside field region. In addition, the XV film had a lower linear dose response range than the EDR film, the efficacy of each film type depends on dose range, the XV being used predominantly for planar dose maps and EDR for combined axial dose maps. High dose lines (matchline effect) were studied with film measurement. Matchlines were caused by a contribution of extra penumbral dose from MLC transmission due to curved leaf ends. An MLC bank leaf stepping program was used with various minor overlap values (0, 0.06, 0.1, 0.14, 0.2 cm) of MLC position. With confirmation by BEAMnrc Monte Carlo simulations, a dosimetric overlap value due to collective effect of scatter and the rounded leaf end transmission equivalent to 0.09 cm leaf overlap was found for a particular weighting of each segment. Note the physical offset value set to avoid leaf collision is an additional 0.05 cm. An overlapping co-incident field technique was used to extend field size, this also showed a small jaw-leaf matchline effect at both edges of an overlap region. An aSi-EPID combined with Varian dosimetry software also showed matchline resolution similar to film. The aSi-EPID, XV film, Pinnacle RTP (version 7.0g and 7.4) and BEAMnrc Monte Carlo were all compared for a 25 segment step and shoot IMRT distribution. IMRT doses in the axial plane were further verified with an I�mRT phantom (Scanditronix-Wellhofer) using the EDR and a new low dose radiochromic film (Gafchromic� EBT, Lot no. 34267-004). For the irradiated perpendicular calibration setup, dose agreed to within &#;5% (1 SD) for EDR and '4% (1 SD) for Gafchromic' EBT film with RTP and an ionization chamber. The conclusions based on this thesis are the following; The matchlines represented a potential overdose to some small volumes within the target dose delivery. The matchline patterns produced by moving leaf banks in known sequences helped reveal the physics properties of the rounded leaf end. Appropriate physical leaf gaps were found to mask the matchline, however due to differences in segment weights these were not recommended. A Monte Carlo model of the Varian 120 MLC was developed using Beam NRC and this model predicted matchline effects. EPID dosimetry revealed an a-Si detector array had sufficient spacial resolution to show matchlines. Late in cycle Version 7.4 of RTP computer leaf model did predict matchlines of smaller magnitude than experimental results.

02Whole.pdf (2503 kB)