ESTRO 2024 - Abstract Book

S4812

Physics - Quality assurance and auditing

ESTRO 2024

rate, 5%/2mm, 20% threshold). The maximum relative movement error for the MLC pairs was noted, along with any visible hot/cold streaks on the difference of plan-to-logfile fluence planes.

Results:

During the recalculation of the clinical patient plans it was found that when using the 5%/2mm criteria almost all the plans had a gamma pass rate of 100%. Therefore, in order to observe small variations in the recalculation, the gamma criteria was changed to 3%/2mm. The gamma pass rate for each individual PTV was also analysed so that differences in the dose distributions around small PTVs would be found even in plans with multiple PTVs. Through analysis of the recalculation results the tolerances were set as shown in table 1. This would allow for the desired 5% failure rate.

For the logging software, the log files were analysed from the deliveries of the PSQA of patient plans during the commissioning process. Seven plans (12%), from four patients, did not get picked up by the software. During testing we found that this was a limitation of Prosoma Core. If there is a patient with multiple plans, at least 10 minutes should elapse between the end of the delivery of one plan and before the initiation of the next. As a direct comparison of the PSQA and the Prosoma Core results was available, it was possible to set the tolerances so that the failures would represent plans that would need review at PSQA as well. Consequently, it was decided that using prominent hot/cold streaks on the difference of plan-to-logfile fluence plane in combination with a MLC relative movement error of more than 2.5% in the same beam would be the markers for failure. The plans that failed the Prosoma Core logging would require to have PSQA done as per standard protocol. Examples of the failures can be seen in Image 1.