Click here to


Are you sure ?

Yes, do it No, cancel

Evaluation of Off-Axis Spatial Accuracy Using An Integrated Quality Assurance Phantom for Frameless Single-Isocenter Multitarget Stereotactic Radiosurgery

D Capaldi1*, L Skinner1 , P Dubrowski1 , A Yu1 , (1) Stanford University Cancer Center, Stanford, CA


(Wednesday, 7/17/2019) 7:30 AM - 9:30 AM

Room: 303

Purpose: The off-axis accuracy of the linear accelerator is crucial for single-isocenter multitarget stereotactic-radiosurgery (SRS). The common Winston-Lutz test is limited to only evaluating accuracy at the isocenter. In this study, a novel quality-assurance (QA) phantom was developed and validated to evaluate off-axis accuracy (i.e. off-axis Winston-Lutz test) to facilitate implementation of a frameless single-isocenter, multitarget SRS program.

Methods: A row of off-axis ball-bearings (BBs) was designed and integrated into a novel 3D printed QA phantom with the ability to QA various positioning systems as well as quantifying dosimetric accuracy. A CT scan of the phantom was acquired, digitally reconstructed radiographic images were generated, and portal images were delivered on a Varian Truebeam with a high-definition multi-leaf-collimator (MLC). To quantify the spatial-accuracy versus distance from isocenter, 2D displacements were calculated between the planned and delivered BB locations relative to their respective MLC defined field boarders. The results for the central BB was validated against commercial software. Furthermore, the phantom was physically shifted by known amounts (0mm, 0.5mm, and 1.0mm) and images were analyzed to determine whether the shifts were identified. Univariate-regression analysis (Pearson correlation coefficient [r]) was performed to quantify relationships.

Results: The proposed QA phantom identified a reduction in spatial-accuracy further away from isocenter. Differences increased as distance from isocenter increased (slope=0.01±0.17;r=0.90;p=0.03) exceeding recommended SRS accuracy tolerances at 7cm away from isocenter. To validate, the average difference between our analysis and commercial Winston-Lutz software was less than 0.02mm for the central BB. Additionally, known shifts of the phantom were accurately identified (slope=1.10±0.13).

Conclusion: The spatial accuracy of a linear accelerator versus distance from isocenter was evaluated using a novel integrated QA phantom. With the ability to quantify off-axis spatial-discrepancies, we can determine limitations on the maximum distance between targets to ensure a single-isocenter multitarget SRS program meets recommended guidelines.


Quality Assurance, Stereotactic Radiosurgery, 3D


Not Applicable / None Entered.

Contact Email