An algorithm for the calculation of dose distributions from dynamic wedges

The Clinac 2100C of Varian has the option of a dynamic wedge. The control of the dynamic wedge is done by means of a segmentation table, which specifies a number of asymmetric fields with corresponding numbers of monitor units. Advantages of this dynamic wedge are the possibility to have more wedge angles and an increased flexibility of operation; a disadvantage is that the measurement of the data needed for a planning system may be time consuming. For the calculation of the dose distribution, we have developed an algorithm which eradicates this disadvantage because it only uses measured data from square fields of the open beam: central axis depth-doses, profiles and Peak Scatter Factors. It is based on a previously developed method which separates the dose into three factors: depth-dose, boundary distribution and envelope profile, and computes the dose of irregular field as well. Depth-dose and boundary distribution are computed by convolution of a field intensity with 'scatter' and 'boundary' pencil beam kernels respectively. For a rectangular field, the field intensity is computed by a weighted sum of intensity matrices, derived from the segmentation table. The matrix elements are equal to 1 within the asymmetric fields and equal to an effective transmission factor under the jaws. The weighting factors in the sum take into account the number of monitor units and the collimator scatter back to the monitor. For an irregular field, the blocked parts of the field are modified, taking into account the transmission through the blocks. This algorithm has been implemented in CADPLAN (Varian-Dosetek) and has been compared with measurements. The results are in good agreement with the international requirements on dose accuracy