Electron dose calculations: a comparison of two commercial treatment planning computers

The accuracy of electron dose calculations performed by two commercially available treatment planning systems, Varian Cadplan and MDS Nordion Helax-TMS, were assessed. Three tests designed to reproduce clinical treatments likely to result in dose nonuniformity have been carried out. The tests examined oblique incidence of the electron beam; incidence on a surface containing a step shape; and incidence on a phantom containing a small air cavity. Dose calculations performed by the planning systems were compared with thermoluminescence dosimetry (TLD) measurements in a WTe electron solid water phantom. A Varian 2100C linear accelerator was used. In most situations, the discrepancy between calculated and measured dose was within the tolerance specified by the ICRU; however, some exceptions were noted. Helax-TMS produced errors of 5 mm in the position of the 10% isodose line in the penumbra of the obliquely incident beam. Both Cadplan and Helax-TMS overestimated the surface dose adjacent to a step in the beam entry surface by approximately 15%. An overestimation of 10% in dose was calculated by both systems downstream of the small air cavity. Discrepancies between the measured and calculated monitor units lay within the uncertainty limits of the measurements. In conclusion, calculations of absorbed dose from electron beams performed by Varian Cadplan and MDS Nordion Helax-TMS result in significant errors at shallow depths near surface irregularities and downstream of small air cavities