Clinical implementation of wedge filter optimization in three-dimensional radiotherapy treatment planning

Background and purpose: To describe a wedge filter optimization technique which automatically chooses the beam weights and wedge filters and to demonstrate the implementation of the algorithm in clinical three-dimensional (3D) radiotherapy treatment planning. Material and methods: Given the incident directions and beam energies of J beams, the dose distribution is a function of the beam weights, wedge angles, and wedge orientations. Instead of decomposing an incident field into a superposition of an open and two nominal wedged fields and then optimizing their weights, the algorithm optimizes the objective function with respect to the beam weights, wedge angles and wedge orientations directly. A salient feature of the algorithm is that no planner intervention was required in the selection of wedge filters during the optimization process. A dose-based objective function which incorporated the relative importance of structures was adopted in this work. The objective function was minimized by the method of simulated annealing. The technique was demonstrated by using a phantom study and two clinical cases. Results: For the phantom case, the classical wedge pair result was obtained, providing a useful test of the algorithm. Dose distributions and dose-volume histograms for the target and surrounding organs were presented for the two clinical cases. It was also shown that dose homogeneity to the target could be compromised by increasing the relative importance factors to the surrounding organs. Conclusions: A 3D wedge filter optimization algorithm has been developed. The technique has the potential to fully automate the 3D radiotherapy treatment planning process. In addition, treatment planning time and efforts were significantly reduced. (C) 1999 Elsevier Science Ireland Ltd. All rights reserved.