Unencrypted login | home

Program Information

Quantification and Prediction of Dose Variation in Scanning Beam Proton Therapy From Setup and Anatomical Change for Single-Field, Multi-Field, and Robustly Optimized Plans


H Li

H Li1*, P Park1, X Zhang1, W Liu1, Z Liao1, L Dong2, M Gillin2, X Zhu2, (1) The University of Texas MD Anderson, Houston, TX, (2) Scripps Proton Therapy Center, San Diego, CA.

TU-A-108-5 Tuesday 8:00AM - 9:55AM Room: 108

Purpose:
The purpose of this study is two-fold: 1) to study proton dose variation due to weekly patient setup and anatomy change; and 2) to evaluate the effectiveness of robustness optimization and robustness evaluation techniques.

Methods:
For a given patient, single-field optimized (SFO), multi-field optimized (MFO) and multi-field optimized with robustness constrains (MFO-Robust) plans were generated on a simulation CT (CT0). The robustness of the plans was evaluated on CT0 with a statistical technique to compute standard deviation of the dose-volume histogram (SD-DVH) of the target volumes and critical structures for 600 combinations of setup and range uncertainties. The original plans were recalculated on subsequent weekly CTs (CTn), and the dose distributions on weekly CTs were deformed onto CT0 using the Velocity (Velocity Medical Solutions) software. The DVHs of weekly doses, as well as the accumulated dose were compared on CT0 with the anticipated SD-DVH variations predicted by the statistical plan robustness evaluation system.

Results:
The mean standard deviation of weekly doses for voxels within the GTV is <2%, for all techniques, the standard deviation of dose difference between the accumulated dose and the planned dose is <1% in GTV for all techniques. The DVH deviations of weekly and accumulated doses were within the 2SD-DVH confidence interval.

Conclusion:
This study quantifies the weekly variation of the patient dose, and the accumulated dose variation due to setup uncertainty and anatomy change. The SD-DVH, which is calculated at the time of planning, could be a good predictor of dose variation over the course of treatment.

Contact Email: