Click here to


Are you sure ?

Yes, do it No, cancel

The Computation and Optimization On IROC Lung Proton Phantom Using Monte Carlo (MC) Algorithm

J Syh*, H Wu , P Nebah , K Chen , Willis-Knighton Medical Center, Shreveport, LA


(Sunday, 7/14/2019)  

Room: ePoster Forums

Purpose: In order to improve monte carlo (MC) based dose computation time and maintain plan optimization, a proton pencil beam plan setting can be determined and optimized to limit the lengthy hybrid computing process. Using IROC lung phantom in MC computation algorithm, the grid size, grid volume and total particle counts can be managed to reduce computation time without compromise the results of an optimal treatment plan. The TLD results were reported.

Methods: A proton treatment plan system has been commissioned to enable MC based calculation, various grid sizes, particle counts, and grid volume of total number of voxels were employed. It was compelling that the optimized number of ion counts be determined first. The effects on grid size or grid volume can be compared independently after. The results of grid size-time was normalized at standard grid size of 2mm. Firstly, at a constant grid volume, total time was compared at various grid size which was increased by 1mm increment. Secondly, different grid volumes were ran with fixed 2mm grid size, timing was verified with grid volume efficiency. The upper and lower TLD capsules were computed separately.

Results: Both results were shown in the attached Figures and Tables.MC based calculations provide as a valuable engine for proton dose calculations with time consuming compared to pencil-beam based due to grid size, grid volume primarily. The hot or cold spots were sparsely found when grid size was greater than 3mm. In this case, 2mm was applied in IROC lung phantom optimization. The IROC TLD outcome of average reading met our MC MC computation results within 1%.

Conclusion: Our work is to optimize plan quality by setting beam computational grid size/volume threshold for MC dose calculations. The significant decline of lengthy computation time in managing grid dimensions with optimal plan is achievable.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email