Purpose: Commission baseline values of the multi-leaf collimator (MLC) using Varian's dynalog file viewer software (DFV) and establish tolerance levels to use as a periodic quality assurance tool in intensity modulated radiotherapy (IMRT).
Methods: A sliding window test was created to have all leafs moving at a constant high speed of 2.5cm/s. This test was performed for 100 days at gantry angles of 0°, 90° and 270° during the morning quality checks in an accelerator performing RapidArc treatments with a Millennium 120 MLC. The DFV was used to collect the percentage of counts in bins 1 (0.05mm) to 8 (deviations in steps of 0.5mm) and the root mean square error (RMS). These were analyzed statistically and a baseline value and tolerance levels were established.
Results: Results show that the greater percentage of counts is found at bin 4 with an average of 97%. Bins 1 to 3 have less than 1 % and bin 5 presents an average of 1.7%. As for bin 6, the average was below 0.1%. Bins 7 and 8 presented no counts. When comparing the results of the different gantry angles, significant differences are only found in bin 4 and 5, with a better performance for the 0° gantry. As for the RMS error, no difference was found between different carriages or gantry angles and the average value was 0.129. A tolerance level was established for bins that correspond to greater deviations - bins 5 to 8, with an upper limit equivalent to twice the standard deviation of each bin.
Conclusions: With the introduction of IMRT, the MLC quality assurance became more extensive and thorough. However a quantitative analysis is often difficult to interpret and time consuming. The purposed method allows the monitorization of the stability and performance of the MLC in a quantitative and easy way.
Keywords: Collimators; Intensity modulated radiation therapy; Multileaf collimators; Quality assurance; Statistical analysis.
© 2012 American Association of Physicists in Medicine.