Radiation Physics

**PV QA 3 - Poster Viewing Q&A 3**

Tuesday, October 23

1:00 PM - 2:30 PM

Location: Innovation Hub, Exhibit Hall 3

**Purpose/Objective(s):**** **To determine the magnitude of dosimetric variations in stereotactic body radiation therapy (SBRT) with respect to radiation isocenter positional errors measured with the Winston-Lutz method. We hypothesize that a correlation between maximum dose errors and radiation isocenter instabilities can be determined to predict the maximum error in the dose delivered to a malignancy in SBRT treatments.

**Materials/Methods:**** **We have performed a retrospective analysis of dosimetric variations due to mechanical instabilities such as shifts of the radiation isocenter caused by gantry sag and couch isocenter walkout during patient set up for non-coplanar beams. We analyzed the dose distributions of 20 patients using a commercial treatment planning system (TPS). We simulated the expected dose differences by shifting the treatment isocenter in the TPS to emulate the daily 3D vector shift of the radiation isocenter detected by the W-L test. The dose errors were classified according to the W-L detected shift magnitude and planning target volume (PTV). Simulated shifts were limited to a maximum of 2.0 mm from the ideal isocenter according to the maximum deviations measured during two consecutive years of W-L data on a SBRT calibrated VERSA HD. A family of hyperbolic curves based on an in-house dose error model of a spherical or ellipsoidal tumor was used to investigate the parametrization of dosimetric variations in the treated malignancy. No stratification between treatment technique and tumor site was performed for our analysis.

**Results:**** **In the considered sample, the dosimetric variations do not follow the expected trend from the spherical or elliptical tumor model enclosing the targeted malignancy. However, the maximum dose errors are bounded by a family of hyperbolic curves that diverge at sample tumor sizes less than 3 cm in diameter and are asymptotic to a minimum error limit for large tumors. Furthermore, for small tumors (< 20 cm^{3}) there are multiple cases where the expected dose error is minimal (<3%) with irradiation isocenter shifts of up to 2.0 mm. No significant difference is shown between shifts oriented at each orthogonal axis.

**Conclusion:**** **Dose variations to the target malignancy due to radiation isocenter perturbations do not follow a parametric curve. However, the maximum dose variation can be modeled using a family of parametric curves where the curve selection is a function of the shift magnitude. This is in part due to high stability cases with respect to radiation isocenter perturbations. The study presents the first steps to determine the factors that improve dose delivery stability to increase dose accuracy in SBRT treatments.

The asset you are trying to access is locked. Please enter your access key to unlock.