Room: AAPM ePoster Library
Purpose: study the scanning path reduction performance for intensity modulated proton therapy based on the improved adaptive genetic algorithm.
Methods: treatment planning system (TPS) used in this study is called DeepPlan which is developed for the self-developed proton therapy facilities. The scanning path optimization process of the three-dimensional IMPT planning is logically similar to the travelling salesman problem (TSP) which is a classic non-deterministic polynomial problem. The scanning path optimization is realized through minimizing the total length equation in each energy slice. The improved adaptive genetic algorithm (IAGA) was chosen to solving the scanning path optimization problem in this study. The test cases adopted were two AAPM TG-119 cases and two corresponding clinical cases. The dose prescriptions were recommended by the AAPM TG-119 report. The scanning path lengths for the four cases using IMPT technique were compared before and after scanning path optimization.
Results: the two TG-119 cases, the results showed that the IAGA optimized scanning path lengths decreased by 32.39% and 14.36% compared to the initial zigzag scanning path. Similar to the TG-119 cases, the initial scanning path lengths decreased by 25.37% and 31.51% after IAGA optimization for the clinical cases. The IAGA optimized scanning path was able to avoid the “hole” area, and significantly reduced the connected paths between the isolated “islands” area, as shown in the supplement.
Conclusion: this study, an IMPT scanning path optimization method was developed based on the improved adaptive genetic algorithm, and was integrated into a self-developed TPS. The AAPM TG-119 cases’ and clinical cases’ test results show that the scanning path optimization module for IMPT can reduce the length of the scanning path thus reducing the IMPT plan delivery time. The intensity modulated heavy-ion therapy scanning path optimization method will be further investigated in the future.