Abstract
PURPOSE: To evaluate the Varian machine performance check (MPC) collimator devices check (CDC) for routine MLC and jaw testing as part of an AAPM compliant linac QA program. METHODS: CDC MLC positioning, MLC backlash, jaw positioning, and jaw parallelism were each assessed for repeatability and concordance with conventional QA. MLC and jaw positioning were also assessed for sensitivity. Measurement time and repeatability of CDC were assessed by timing and recording five successive measurements on a single linac. Concordance was assessed monthly over 5 months on four linacs, conducted during the same session as conventional QA. MLC positioning was compared to an advanced picket fence test, while jaw positioning and parallelism were compared to department in-house EPID based methods. MLC backlash was compared to the Varian built-in method. Sensitivity was assessed via deliberately introduced errors except for MLC backlash, which was assessed via correlation between methods across leaf banks. RESULTS: CDC requires 4:09 (min:s) ± 1.8 s (2 SD) to perform. Repeatability was measured to be: 0.02 mm for both MLC positioning and backlash, 0.15 mm for jaw positioning and 0.009° for jaw parallelism (2 SD). Concordance was observed for mean MLC positioning to within 0.32 , 0.08 mm for MLC backlash, 0.6 mm for jaw positioning and 0.06° for jaw parallelism. MLC and jaw positioning sensitivity were observed with maximum mean difference between methods of 0.18 and 0.71 mm, respectively. MLC backlash correlation coefficient between methods across leaf banks was observed to 0.84 and 0.9 for banks A and B, respectively. CONCLUSION: MPC CDC has been demonstrated to provide acceptably equivalent MLC and jaw positioning assessment to standard methods and could conceivably be used in a linac QA program for these purposes.