Dose verification of IMRT by use of a COMPASS transmission detector.

Our purpose in this study was to evaluate the fundamental accuracy of reconstructed dose distributions from the COMPASS system using specific MLC test patterns and complicated IMRT neck plans. The COMPASS-reconstructed dose distributions were compared with those measured with EPID, MapCHECK, and EDR2 film and as well as Monte Carlo-calculated dose profiles with use of square-wave chart patterns of 20-, 10-, and 5-mm gaps and step and pyramid patterns. Additionally, the COMPASS dose distributions for clinical IMRT neck plans were tested. The COMPASS dose profiles were almost in agreement with the Monte Carlo-calculated dose profiles and point doses measured with MapCHECK for 20- and 10-mm gap patterns. The dose profile for a 5-mm gap pattern showed a narrow width due to the detector size in the penumbra region. For step and pyramid patterns, COMPASS agreed with MapCHECK and Monte Carlo calculation, except for EDR2 film. The COMPASS and MapCHECK dose distributions agreed with that of a treatment planning system by gamma analysis (criteria; 3 mm/3%). In comparisons of clinical IMRT neck dose distributions, COMPASS was measured with almost the same accuracy as MapCHECK, but slight deviations were found for large IMRT fields. These deviations could be minimized by improvement of the beam model of the COMPASS system. The COMPASS system can be expected to be used for traditional QA methods in clinical routine with the same accuracy as a MapCHECK diode detector.