ESTRO 2024 - Abstract Book

S4843

Physics - Quality assurance and auditing

ESTRO 2024

Algorithms analysis accuracy was evaluated by repeating measurements in the same conditions and assessed to be smaller than 0.15 mm for all parameters.

For the WL analysis, AAPM recommends a tolerance of 2 mm diameter. We set a tolerance of 1.5 mm diameter for gantry and couch isocenter, 1 mm diameter for collimator isocenter. Secondarily, we centered the WL phantom with the CBCT and compared results for 6MV and 6MV FFF. This analysis allowed us to correctly match both 6MV and 6MV FFF isocenters with the CBCT isocenter (Figure1 top). For the MLCs analysis, we modified the Pylinac algorithm to calculate absolute leaves positions (as recommended by TG-142) and single leaf error analysis. Absolute leaves positions are calculated from the radiation field center, identified by the use of 4 additional square fields images at different collimator positions. Relative leaves positions to the median leaves position are also performed. We evaluated the algorithm by introducing known errors in single leaves and MLC banks. The modified algorithm was able to detect leaves and bank errors beyond 0.3 mm and correctly identify failing leaves numbers. The use of 4 additional square fields images is essential to correctly identify leaves numbers too. AAPM TG 142 recommends a tolerance of 1 mm, in our protocol we decided to set stricter values with attention level at 0.3 mm and action level at 0.5 mm (Figure1 bottom). As a matter of facts, the new AAPM guidelines(3) suggest as well stricter values for MLC errors. Moreover, the same guidelines underline the importance of a quantitative evaluation of MLC errors with a weekly frequency.