Abstract
We examined the dose distributions generated by Pinnacle3 (Philips Radiation Oncology Systems, Milpitas, CA) for intensity-modulated radiotherapy (IMRT) plans using a cubic-block-piled compensator as the intensity modulator for 4-MV and 10-MV photon beams. The Pinnacle treatment planning system (TPS) uses an algorithm in which only the physical density of the absorber is required for calculating the characteristics of the modulator. The intensity modulator consists of cubic blocks (attenuator) of a tungsten alloy, plus cubic blocks of polyethylene resin foam that fill the spaces between the attenuator blocks and polymethyl methacrylate (PMMA) boards that act as the platform for the modulator. By measuring the transmission for various thicknesses of attenuator and by deriving values for the total physical density of the modulator, we determined the optimal effective density by comparing the curves fitted for the actual transmission data with the transmission calculated by the TPS. Using these effective densities, we examined the accuracy of Pinnacle3 for dose profiles of specific geometric patterns. The levels of consistency between the measurements and the calculations were within a tolerance of 3% of the dose difference and had a 3-mm distance to agreement for the ladder-, stairstep-, and pyramid-shaped test patterns, except in the high dose gradient region. In this modulator assembly, leakage occurred from the slits between the cubic blocks. This leakage was about 1.6% at maximum, and its influence on dose distribution was not crucial. In the TPS, in which physical density was the only usercontrollable parameter, we used the effective density of the absorber deduced from the effective mass attenuation coefficient. We conclude that the intensity modulation compensator system, together with a piled cubic attenuator, is clinically applicable, with an acceptable tolerance level. For the intensity map of the IMRT plan, measurements in treatment fields met 3% and 3-mm criteria, excluding some regions of high gradient, which had a discrepancy of less than 5% and 4 mm.