Plug pattern optimization for gamma knife radiosurgery treatment planning.

PURPOSE: To develop a novel dose optimization algorithm for improving the sparing of critical structures during gamma knife radiosurgery by shaping the plug pattern of each individual shot. METHOD AND MATERIALS: We first use a geometric information (medial axis) aided guided evolutionary simulated annealing (GESA) optimization algorithm to determine the number of shots and isocenter location, size, and weight of each shot. Then we create a plug quality score system that checks the dose contribution to the volume of interest by each plug in the treatment plan. A positive score implies that the corresponding source could be open to improve tumor coverage, whereas a negative score means the source could be blocked for the purpose of sparing normal and critical structures. The plug pattern is then optimized via the GESA algorithm that is integrated with this score system. Weight and position of each shot are also tuned in this procedure. RESULTS: An acoustic tumor case is used to evaluate our algorithm. Compared to the treatment plan generated without plug patterns, adding an optimized plug pattern into the treatment planning process boosts tumor coverage index from 95.1% to 97.2%, reduces RTOG conformity index from 1.279 to 1.167, lowers Paddick's index from 1.34 to 1.20, and trims the critical structure receiving more than 30% maximum dose from 16 mm(3) to 6 mm(3). CONCLUSIONS: Automated GESA-based plug pattern optimization of gamma knife radiosurgery frees the treatment planning team from the manual forward planning procedure and provides an optimal treatment plan.

Duke Scholars

Author Qingrong Wu Radiation Oncology