Room: AAPM ePoster Library
A computer application has been developed to extract and analyze control system generated log files for an Elekta VersaHD linear accelerator (linac).
Geometric and dose parameters are recorded by the linac control system at a sampling rate of 25 Hz during delivery. The application was developed using Python scripting language and utilized the Multileaf Collimator (MLC) recorded data to compute and display leaf speed and position error. MLC-defined sliding window, designed by the Dynamic-MLC (DMLC) mode, was used for leaf speed and position testing.
Log derived results showed the prescribed and actual mean leaf speeds differed by < 1.0% for all leaf speed tests. MLC position errors, mean Room Mean Square (RMS) errors, increased linearly with increasing leaf speed. At maximum leaf speed (35.0 mm/s), the mean RMS error was 0.97± 0.14, while leaf speeds 32.0 and 29.0 mm/s had a mean RMS error of 0.64 ± 0.08 and 0.55 ± 0.06, respectively. The maximum leaf speed can be restricted in a Treatment Planning System (TPS) to achieve an optimal leaf position accuracy.
A log based analytical program can be used as a supplementary tool to better characterize MLC leaf speed parameters to improve the accuracy of DMLC delivery. This tool can also be used to generate quantitative results when performing quality assurance (QA) testing on MLC leaf speed constancy.
Not Applicable / None Entered.