Because of high precision and mild side effects, intensity-modulated proton therapy (IMPT) has become a hot spot in the radiotherapy field. Nevertheless, the precision of IMPT is extremely sensitive to the range uncertainties. In this paper, a novel robust optimization method was proposed to reduce the effect of range uncertainty upon IMPT.

Firstly, the robust optimization model was established which contained three types of range including the increased range, the normal range and the shortened range. The objective function was expressed in quadratic function. The organ dose contribution matrix of each range was calculated by proton pencil beam algorithm. The range deviation was discretized and the probability of each range was obtained based on the Gauss distribution function. Finally, the conjugate gradient method was adopted to find the optimal solution to make the actual dose coverage of the target area and the organs at risk distributed within the expected dose as possible.

The 3 sets of simulation tests provided by the AAPM TG-119 Report were utilized to evaluate the effectiveness of this method: nasopharyngeal carcinoma, prostate and " C" -type cases. Compared with conventional IMPT optimization approach, this novel method was less sensitive to the range uncertainty. When the range deviation occurred, the dose coverage of the target area and organs at risk of the nasopharyngeal carcinoma and prostate cases almost reached the expected dose, and the high dose coverage of the target area and organs at risk protection were improved in the" C" -type cases.

To compensate for the range uncertainty, this novel method can enhance the dose coverage of the target area and reduce the dose coverage of the organs at risk.

Range uncertainty; Intensity-modulated proton therapy; Robust optimization