Encrypted login | home

Program Information

Towards Patient-Specific 4D Quality Control of VMAT Plans for High-Dose-Rate SBRT

J Park

J Park*, F Li , S Samant , B Lu , J Li , C Liu , G Yan , University of Florida, Gainesville, FL


TH-AB-FS1-16 (Thursday, August 3, 2017) 7:30 AM - 9:30 AM Room: Four Seasons 1

Purpose: In SBRT using VMAT, rigorous patient-specific 3D QA is mandatory. However, quality control (QC) concerning the 4D nature is largely overlooked. We developed a flexible 4D dose accumulation (4DDA) framework towards patient-specific 4DQC of VMAT plans for SBRT.

Methods: Unlike existing 4DDA methods that assume repeatable breathing patterns, our framework uses realistic breathing patterns collected with an optical tracking system. The control points (CPs) of VMAT plans were upsampled to 1o angular separation, then temporally aligned to the motion signal. We shifted the isocenters of individual CPs in directions opposite to the motion. The 4D dose was accumulated from CP dose calculated with a commercial treatment planning system on the end-exhale CT. Using the framework, we studied (1) patient-specific dosimetric difference (interplay effect) between the plan and the delivery; (2) dosimetric effect of breathing pattern variations (amplitude, period, baseline drift); (3) effect of starting phase in high-dose-rate SBRT.

Results: We achieved nearly identical dosimetric endpoints before and after the CP manipulation when there is no motion, which validated the accuracy of the 4DDA framework. For VMAT plans with low modulation (single lung target close to rib case), the interplay effect was within 2%; however, if there was significant breathing irregularities, the discrepancy increased to ~5%. For plans with high modulation (multiple targets adjacent to each other or to critical structures), we found dosimetric differences up to 20%. Among breathing pattern variations, baseline drift had the most pronounced dosimetric impact. The dosimetric effect of different beam-starting phase reached 10% in some cases.

Conclusion: Our results suggested that respiratory motion-induced dosimetric difference depends on patient-specific factors (plan complexity, breathing motion variations, etc) which necessitates patient-specific 4DQC. The proposed framework offers us opportunity to take preventive actions in anticipation of adverse outcome.

Contact Email: