Abstract
A comparison of the monitor unit calculations of a commercial 3D computerized treatment planning system (TPS) with "hand" calculations from lookup tables was made for a large number of clinical cases (greater than 13 500 treatment fields). Differences were analyzed by treatment site for prostate, rectum, cranium, and breast. The 3D TPS monitor unit calculation was systematically higher than the "hand" calculation by an amount that depended on the complexity of the treatment geometry. For simple geometries the mean difference was 1% and was as high as 3% for more complicated geometries. The higher value was attributed to an accumulation of differences introduced by multiple factors in the monitor unit calculation. Careful attention to factors such as patient contour could reduce the mean difference. "Hand" calculations were shown to be an accurate and useful tool for verification of TPS monitor unit calculations.