SU‐FF‐J‐05: A Four‐Dimensional Planning and Monte‐Carlo Dose Calculation Framework for Real‐Time Tumor Motion Compensation

Abstract

Purpose: To perform 4D dose calculations using a feedback‐controlled treatment couch in 15 lung and abdominal patients Methods: Fifteen patients undergoing 4D CT were considered in this work. The GTV was delineated on ten 3D CT data sets by independent experts, with a planning margin added corresponding to the 2σ deviation from the mean center‐of‐mass. Tumor trajectories were derived from 4D CT data and fitted to a Fourier series expansion. Intra‐ and inter‐fraction statistical variations in amplitude and period were applied by considering literature estimates of the standard deviations. The couch feedback control system was simulated using second‐order couch dynamics and a first‐order controller. The residual tumor trajectory was calculated by inputting the tumor trajectory to the control system. A 7‐beam IMRT plan was generated on the end‐exhale CT with a prescription dose of 66 Gy. The resulting 4D dose distribution was calculated by sampling the isocenter position along the continuous tumor trajectory before and after motion compensation using a previously published Monte Carlo algorithm. Results: The mean tumor motion amplitude was 11.4±3.4 (SD) mm. After compensation the amplitude was 0.42±0.23 mm. Degradation of the tumor dose up to 35% was observed. After compensation, an average of 18.8 ± 4.7% improvement in the 100% dose coverage was seen for end inhale, and 13.4 ± 4.7% for the 90% coverage. For the combined 4D dose, GTV dose coverage was 94.7±4.7% (for 100% coverage) after compensation, but degraded to 87.3±4.5% without compensation. Conclusion: We have developed a framework for performing 4D dose calculations using real‐time tumor trajectories. For IMRT plans, the average improvement in the dose coverage to the PTV for mobile tumors with a feedback‐controlled couch was X% Conflict of Interest: Supported by 3DLine Medical Systems.