SU-E-T-170: A Viable Approach to Patient Specific QA for Spine VMAT SRS Using EPID-Based Dosimetry.

PURPOSE VMAT involves multi-parameter modulations, thus presenting multi-dimensional challenges. Here, we present a viable approach to VMAT patient specific QA using EPID-based dosimetry for spine VMAT SRS casesMethods: Our spine VMAT SRS plans consisted of two full arcs and were optimized with 6 MV photons for a Varian Trilogy LINAC. The QA plans were computed by resetting all control points to 0° gantry angle while keeping the MLC apertures and corresponding MU weights intact. The dose distribution at the depth of clinical significance was calculated in a solid water phantom. EPIDose (Sun Nuclear Corp) was used to convert MV EPID images into absolute dose in a plane of interest in a homogeneous medium. During measurements, the source-to-detector distance (SDD) was set to 105 cm and the VMAT QA plan was delivered with 600 MU/min. For each QA plan, three measurements were made. The average pass rate (thresholds: 2%/2 mm) was then calculated. In addition, the dose at the isocenter or other points of clinical significance was also measured and the mean was calculated. RESULTS Our current institutional QA acceptance criteria for VMAT plans are: pass rate 90% for 2%/2mm with a dose threshold of 10% and the discrepancy between the measured and planned doses at the isocenter or other points of clinical significance 2%. Our preliminary investigation indicated that the pass rate was normally greater than 95.0% with a total number of valid detection points of ∼30000. The dose difference at the point of interest was, in general, around ±1.0 %. Sharp dose gradients were accurately detected at the PTV-cord interface. CONCLUSIONS EPID- based dosimetry is real-time and financially viable. It can achieve sub- millimeter dosimetry accuracy without extensive inter-diode interpolations. We believe that EPID-based VMAT dosimetry offers a competitive alternative to other competing dosimetry technologies.