ePoster Forums
Purpose: To introduce a new concept – the difference of pseudo fluences (DPF) – for fast plan adaption to inter-fractional organ deformations in VMAT plans without the need to re-optimize; to demonstrate the feasibility of the proposed method.
Methods: While computed tomography reconstructs a 3D density distribution from 2D Radon projections via filtered backprojection, radiation therapy aims to generate 2D fluence data from a desired 3D dose distribution. This inverse relation motivates the definition of the fluence-like “pseudo fluence” as the filtered Radon transform of a given 3D dose distribution. When changes in the dose distribution are required due to anatomical deformations, calculating the difference of this projection of a desired dose and of a deformed distribution – the DPF – is expected to indicate areas for leaf adaption.To verify this claim, the pseudo fluence difference was calculated for a phantom and compared with the monitor unit (MU)-weighted leaf position changes in plans further optimized for inter-fractionally deformed structures in a commercial treatment planning system.
Results: Comparing the difference of pseudo fluences and the MU-weighted leaf position difference shows a strong correlation for most control points: Positive (negative) DPF-peaks frequently coincide with areas, where leafs are opened (closed).
Conclusion: This work establishes the potential of the difference of pseudo fluences as a new measure for leaf adaption purposes and demonstrates its significant similarity to the leaf changes made by the treatment planning system. It demonstrates that local maxima and minima in the DPF are used as an indicator, where leafs need to be opened or closed, respectively, in order to adapt the fluence. Further work will quantify the correlation between the pseudo fluence difference and leaf changes. In future investigations, pseudo fluence based predictions of necessary leaf shifts will be shown.
Funding Support, Disclosures, and Conflict of Interest: This work was in part funded by research grants of Wilhelm-Sander-Stiftung SANST # 2018.089.1 (until June 2021 - development of basic ideas) and Elekta AB (Stockholm, Sweden) (since July 2021 - further development).