THE RELIABILITY OF OPTIMIZATION UNDER DOSE-VOLUME LIMITS

Purpose: An optimization algorithm improves the distribution of dose a mong discrete points in tissues, but tolerance depends on the distribu tion of dose across a continuous volume. This report asks whether an e xact algorithm can be completed when enough points are taken to accura tely model a dose-volume constraint. Methods and Materials: Trials wer e performed using a 3-dimensional model of conformal therapy of lung c ancer. Trials were repeated with different limits placed on the fracti on of lung which could receive > 20 Gy. Bounds were placed on cord dos e and target dose inhomogeneity. A mixed integer algorithm was used to find a feasible set of beam weights which would maximize tumor dose. Tests of feasibility and optimality are introduced to check the soluti on accuracy. Results: Solutions were optimal for points used to model tissues. An accuracy of 3-4% in a volume condition could be obtained w ith models of 450-600 points. The error improved to 2% with 800 points to model the lung. Solution times increased six-fold at this level of accuracy. Conclusion: The mixed integer method can find optimum weigh ts which respect dose-volume conditions in usually acceptable times. I f constraints are violated by an excessive amount, the optimization mo del should be rerun with more points.