GENERIC GENETIC ALGORITHM FOR GENERATING BEAM WEIGHTS

A genetic algorithm for generating beam weights is described. The algo rithm improves an objective measure of the dose distribution while res pecting dose volume constraints placed on critical structures. The alg orithm was used to select beam weights for treatment of abdominal tumo rs. Weights were selected for up to 36 beams. Dose volume limits were placed on normal organs and a dose inhomogeneity limit was placed on t umor. Volumes were represented as sets of several hundred discrete poi nts. The algorithm searched for the beam weights that would make the m inimum tumor dose as high as the constraints would allow. The results were checked using dose volume histograms with standard sized grids. N ineteen trials were created using six patient cases by changing the re quired field margin or allowed beam position in each case. The samplin g of points was sufficiently dense to yield solutions that strictly sa tisfied the constraints when the prescribed dose was renormalized by a factor of less than 6%. The genetic algorithm supplied solutions in 4 9 min on average, and in a maximum time of 87 min. The randomized sear ch does not guarantee optimality, but high tumor doses were obtained. An example is shown for which the solution of the genetic algorithm ga ve a minimum tumor dose 7 Gy higher than the solution given by a simul ated annealing algorithm under the same set of constraints. The geneti c algorithm can be generalized to admit nonlinear functions of the bea m intensities in the objective or in the constraints. These can includ e tumor control and normal tissue complication probabilities. The gene tic algorithm is an attractive procedure for assigning beam weights in multifield plans. It improves the dose distribution while respecting specified rules for tissue tolerance. (C) 1996 American Association of Physicists in Medicine.