(Sunday, 7/29/2018) 4:00 PM - 4:30 PM
Room: Exhibit Hall | Forum 7
Purpose: To improve gross error detectability of EPID-based real-time cine frame-by-frame during-VMAT treatment delivery verification by eliminating false positive errors detected at the dynamic beam aperture.
Methods: In cine exit-fluence verification, measured frames are compared with those predicted through patient planning CT. Monitor units (MUs) are used to synchronize measured-predicted frame pairs. A buffered masking technique is used to verify the aperture, ensuring that radiation exists within the aperture-buffer and radiation is absent outside of the aperture+buffer. A uniform 2 mm buffer, independent of MLC leaf velocity, is compared with an MLC-leaf specific mask which considers the MLC-leaf positions at +/- 1 MU with a 1 mm buffer. A frame is tagged as a gross delivery error if a contiguous 64 mm^2 area of extra/missed radiation is detected and a warning at 40 mm^2. Detectability is evaluated for 23 during-treatment VMAT deliveries, resulting in ~12000 EPID frames.
Results: Compared with the uniform 2 mm buffer, the dynamic buffer decreased gross error reports from 9.1% to 0% and warnings from 28.3% to 23.4%. The integral area of the dynamic buffer is equivalent (within <1%) to that of the uniform buffer.
Conclusion: Frame-by-frame EPID-based verification is challenging by MLC leaf motion during frame acquisition and MLC leaf positioning uncertainties. Direct checking of the aperture, with a buffer which accounts for delivery-specific MLC leaf motion effectively detects aperture variations while eliminating false positive gross errors related to the beam aperture. 
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email
