Author: Qihui Lyu, Dan Ruan, Ke Sheng, Jingjie Yu 👨🔬
Affiliation: Department of Radiation Oncology, University of California, Los Angeles, University of California, San Francisco, Department of Radiation Oncology, University of California, San Francisco 🌍
Purpose: The robotic arm radiotherapy platform enables flexible delivery of non-coplanar and non-isocentric radiotherapy with variable Source-to-Isocenter Distances (SIDs). However, the high degrees of freedom in treatment planning intensify the tradeoff between plan quality and beam efficiency. This study evaluates the use of multiple SIDs to achieve both high plan quality and efficiency in robotic radiotherapy.
Methods: We developed an optimization framework that integrates beam orientation optimization (BOO), isocenter selection, SID selection, and fluence map optimization (FMO). The algorithm includes a least-square fidelity objective for dosimetry, a total variation term to reduce the number of fluence segments, and a group sparsity term for selecting optimal beams from a candidate beam pool with different angles, SIDs, and isocenters. Multi-SID intensity-modulated radiation therapy (IMRT) plans were compared with fixed-SID 4π IMRT plans (SID-50 and SID-100) across 10 head-and-neck (H&N) cancer cases. We also ran beam sensitivity test and created plans with 10 to 30 beams to evaluate the effect of beam number.
Results: The datasets included 28 PTV regions across 10 patients. The 15-beam Multi-SID plan demonstrated superior PTV homogeneity compared to the fixed-SID plans (SID-100 and SID-50). The Multi-SID plan achieved the lowest D2 in 13 PTV regions and the highest D98 in 16 PTV regions. Specifically, the average D2 values were 1.140±0.074 for Multi-SID, 1.147±0.074 for SID-100, and 1.149±0.074 for SID-50. For D98, the average values were 0.976±0.022 for Multi-SID, 0.975±0.023 for SID-100, and 0.974±0.022 for SID-50. For beam sensitivity test, 10-beam multi-SID and SID-100 plans are significantly better than 10-beam SID-50 plans. For plans with greater than 25 beams, multi-SID and SID-50 achieved better dosimetry than SID-100.
Conclusion: The proposed multiple SID optimization achieved more desirable dosimetry and/or more efficient treatment plans than fixed SID methods.