Fluence Aperture Alternating Optimization (FAAO) for HYPERSCAN Adaptive Aperture Treatment Planning
s zhou*, Q Chen, Y Tan, T Mazur, T Zhao, R Khan, T Zhang, Washington University School of Medicine, St. Louis, MO
Presentations
SU-E-206-5 (Sunday, 7/10/2022) 1:00 PM - 2:00 PM [Eastern Time (GMT-4)]
Room 206
Purpose: Mevion’s HyperScan system is equipped with an adaptive aperture (dynamic MLC) that can trim off the spots and produce beam penumbra sharper than pencil beam scanning (PBS) beams. We developed a novel treatment planning method that can take advantage of the adaptive aperture by alternatively optimization of aperture and fluence.
Methods: Different from the 2D fluence maps of x-ray intensity-modulated radiotherapy (IMRT), the fluence maps of proton beams represent the weights of ideal monoenergetic pencil beams and are three-dimensional. The inverse optimization algorithm optimizes the 3D fluence maps while keeping the fluence maps smooth in 3D. The optimized 3D fluences are converted to spots and apertures at different energy layers by minimizing the difference between optimized and converted fluences. After conversion, the weights of each spot are optimized again to further improve the plan quality.
Results: The energy spectrum and intensity profiles of each energy layers were obtained by decomposing spot dose distributions into the summation of ideal monoenergetic proton beams. Both energy spectrum and intensity profile can be approximately represented by a double-Gaussian function. Inverse optimization produced smooth 3D fluences with only sharp gradient at the boundary, which was converted to a deliverable plan that comprises spots and MLC positions at optimized energy levels
Conclusion: By alternatively optimizing both aperture and intensities, the FAAO planning method can take full advantage of the sharp beam dose gradient generated by MLC and produce a better plan quality than that conventional planning method that only optimizes spots weights.
Funding Support, Disclosures, and Conflict of Interest: Project supported by Mevion Medical Systems
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email