Comparison of simulated annealing algorithms for conformal therapy treatment planning

摘要: Purpose : The efficiency of four fast simulated annealing algorithms for optimizing conformal radiation therapy treatment plans was studied and the resulting were compared with each other to optimized conventional plans. Methods Materials Four selected on basis their reported successes in solving minimization problems a Cauchy generating function, Lorentzian variable step size generalized (VSGSA), very reannealing (VFSR). They tested six clinical cases using multiple beam coplanar technique. Relative weights computed that maximized minimum tumor dose subject dose-volume constraints normal organ doses. Following some initial tuning parameters, algorithm applied identically test case. Optimization tests run different random number sequences numbers iterations. Results VSGSA consistently produced best results. Using long times, it generated highest five cases. For short solutions averaged larger doses than those all patients, increases ranging from 0.4 5.9 Gy. three plan gave clinically significant increase over plan, 8.2 13.0 In two cases, there little difference between approaches. one case, much better because arrangement included wedges, which offset advantage Conclusions equal computing times both duration, superior by algorithms. simple technique used this study showed potential abdominal tumors. important