Program Information
Intra-Fraction Proton Beam-Range Verification with PET Imaging: Feasibility Studies with Monte Carlo Simulations and Statistical Modeling
K Lou1,2*, D Mirkovic1 , X Sun1 , X Zhu1 , F Poenisch1 , D Grosshans1 , J Clark2 , Y Shao1 (1) U.T M.D. Anderson Cancer Center, Houston TX,(2) Rice University, Houston, TX
Presentations
SU-E-J-82 Sunday 3:00PM - 6:00PM Room: Exhibit HallPurpose:To study the feasibility of intra-fraction proton beam-range verification with PET imaging.
Methods:Two phantoms homogeneous cylindrical PMMA phantoms (290 mm axial length, 38 mm and 200 mm diameter respectively) were studied using PET imaging: a small phantom using a mouse-sized PET (61 mm diameter field of view (FOV)) and a larger phantom using a human brain-sized PET (300 mm FOV). Monte Carlo (MC) simulations (MCNPX and GATE) were used to simulate 179.2 MeV proton pencil beams irradiating the two phantoms and be imaged by the two PET systems. A total of 50 simulations were conducted to generate 50 positron activity distributions and correspondingly 50 measured activity-ranges. The accuracy and precision of these activity-ranges were calculated under different conditions (including count statistics and other factors, such as crystal cross-section). Separate from the MC simulations, an activity distribution measured from a simulated PET image was modeled as a noiseless positron activity distribution corrupted by Poisson counting noise. The results from these two approaches were compared to assess the impact of count statistics on the accuracy and precision of activity-range calculations.
Results:MC Simulations show that the accuracy and precision of an activity-range are dominated by the number (N) of coincidence events of the reconstructed image. They are improved in a manner that is inversely proportional to 1/sqrt(N), which can be understood from the statistical modeling. MC simulations also indicate that the coincidence events acquired within the first 60 seconds with 10^9 protons (small phantom) and 10^10 protons (large phantom) are sufficient to achieve both sub-millimeter accuracy and precision.
Conclusion:Under the current MC simulation conditions, the initial study indicates that the accuracy and precision of beam-range verification are dominated by count statistics, and intra-fraction PET image-based beam-range verification is feasible.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by a research award RP120326 from Cancer Prevention and Research Institute of Texas.
Contact Email: