Ballroom C
Purpose: Electron beam has been commonly utilized for ultra-high dose rate(FLASH) studies and LINAC has been adopted and converted for FLASH research. We are developing a Monte-Carlo(MC)-based treatment planning system(TPS) combined with an in-house GPU-based dose planning method(gDPM) engine for fast dose calculation. To further accelerate dose calculations for animal study, we utilize pre-calculated phase-space for commonly used beam geometry and apply it for gDPM computation.
Methods: The head geometry of Varian 2100EX, the LINAC undergoing modification for FLASH research in our center, was explicitly modeled in Geant4-based GAMOS MC package. The beam model for conventional 18MeV electron beam was established and tuned to fit Varian Golden Beam Data(GBD). For the beam model tuning, we adjusted the parameters of the primary electron beam, including mean energy(Ε), mean energy spread(σₑ), spot size(σ) and mean angular spread(θ). We obtained the FLASH beam model by removing scattering foil in the head geometry model while other components remain, consistent with the procedure of converting our LINAC to FLASH mode. To achieve fast dose calculation necessary for animal research, we pre-calculated the phase-space at SSD=95cm using GAMOS and used it as the beam source for the gDPM computation. The planning studies for animals will be carried on based on the FLASH MC-gDPM platform.
Results: The uncollimated PDD of the conventional 18MeV beam model agreed with GBD at the absolute difference within 1% at Ε=19.45MeV, σₑ=1.0MeV, σ=0.7mm, and θ=0.5°. Compared to the conventional 18E beam, the FLASH beam contains less bremsstrahlung in PDD and shows an immediate decrease in dose away from the central axis in the in-air profile due to the lack of scattering foil.
Conclusion: The success of this work will provide investigators an open-source platform for fast FLASH dose calculation, which can be readily expanded to other LINAC configurations to facilitate FLASH research.
Funding Support, Disclosures, and Conflict of Interest: CPRIT RR200042 (no NIH funding)