Click here to


Are you sure ?

Yes, do it No, cancel

Evaluating Machine Performance in Off-Isocenter Coincidence Using Automatically Acquired Orthogonal KV and MV Pairs

R Patel*, H Kang , Loyola University Medical Center, Maywood, IL


(Sunday, 7/29/2018) 5:05 PM - 6:00 PM

Room: Karl Dean Ballroom A2

Purpose: For multi-lesion SRS treatments with a single isocenter, delivery errors can be magnified compared to multiple isocenter delivery. Consequently, evaluating coincidence of the mechanical and radiation field at off-isocenter becomes crucial. We propose a robust method to evaluate off-isocenter coincidence using a commercial phantom.

Methods: Orthogonal MV/kV pairs of the Optical Surface Monitoring System (OSMS, Varian, CA) isoCube phantom at 18 gantry angles were simultaneously and automatically acquired in the TrueBeam Developer mode. The phantom was manually aligned close to the OSMS isocenter, and then zeroed out via couch shifts. The phantom consists of a central BB and 4 additional BB's offset between 1 –3 cm in 3 dimensions (3D). The locations of the 5 BB's were determined in the MV and kV images using a Matlab (Mathwork, MA) algorithm. The algorithm analyzes the locations of the BBs in each MV/kV pair, and then outputs deviations between the expected and measured positions. In-plane BB offsets were measured in the MV images, and out-of-plane BB offsets relative to MV image plane were measured in the kV images. This process was repeated over 3 consecutive weeks.

Results: The results using our stereotactic linac (Varian Edge) showed that the mean 3D displacement for the central BB was 0.45±0.12mm with a maximum of 0.66mm. The mean and standard deviation of the displacements of the offset BB's were between 0.48±0.15mm – 0.57±0.26mm with a maximum of 1.13mm for the most offset BB.

Conclusion: We developed an off-isocenter Winston-Lutz test to evaluate the accuracy of single iscoenter delivery for multi-lesions SRS. This method assesses displacement in 3D at isocenter and off-isocenter locations using orthogonal images whereas conventional Winston-Lutz test only evaluates in 2D at isocenter. This test can be implemented to verify the consistency of linac performance before treating multi-lesion SRS with a single isocenter.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email