Ein Konzept zur Optimierung der Strahlentherapie

Abstract

A method for the optimisation of intensity modulated radiotherapy (IMRT) with an emphasis on clinical and biological aspects is presented. Various models for the probability of cell survival as a function of single fraction and total dose form the objective function for the tumour tissue, which is to be minimized. The biological dose-response in normal tissues at micro-, meso- and macroscopic length scales represent constraints to the optimum. Both objective function and constraints are functionals of the dose distribution which can be reduced to effect densities by a mean-field approximation. The total effect is the integral of these densities over the organ volume. The optimum dose in the target volume results as the maximum under strict adherence to the normal tissue constraints. The fluence profiles are subject to a smoothing operation which ensures treatment efficiency. The constrained variational problem can be reduced to a finite optimisation problem by introducing a basis in fluence space. A phenomenological dose computation algorithm is presented which is designed for this fluence basis. This algorithm is merged with a Monte Carlo dose computation algorithm to form a hybrid which is capable of computations on clinical time scales. The combination of IMRT and Monte Carlo makes it possible to compensate for electron scatter effects at surfaces by modulating the primary fluence. The present concept explores the potential of IMRT for better chances for cure, which is demonstrated with a number of clinical examples.

A method for the optimisation of intensity modulated radiotherapy (IMRT) with an emphasis on clinical and biological aspects is presented. Various models for the probability of cell survival as a function of single fraction and total dose form the objective function for the tumour tissue, which is to be minimized. The biological dose-response in normal tissues at micro-, meso- and macroscopic length scales represent constraints to the optimum. Both objective function and constraints are functionals of the dose distribution which can be reduced to effect densities by a mean-field approximation. The total effect is the integral of these densities over the organ volume. The optimum dose in the target volume results as the maximum under strict adherence to the normal tissue constraints. The fluence profiles are subject to a smoothing operation which ensures treatment efficiency. The constrained variational problem can be reduced to a finite optimisation problem by introducing a basis in fluence space. A phenomenological dose computation algorithm is presented which is designed for this fluence basis. This algorithm is merged with a Monte Carlo dose computation algorithm to form a hybrid which is capable of computations on clinical time scales. The combination of IMRT and Monte Carlo makes it possible to compensate for electron scatter effects at surfaces by modulating the primary fluence. The present concept explores the potential of IMRT for better chances for cure, which is demonstrated with a number of clinical examples.