Monte Carlo evaluation of the dosimetric uncertainty in matched 6 MV Elekta and Varian linear accelerators

Lye, J 2012, Monte Carlo evaluation of the dosimetric uncertainty in matched 6 MV Elekta and Varian linear accelerators, Masters by Research, Applied Sciences, RMIT University.


Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Lye.pdf Thesis Click to show the corresponding preview/stream application/pdf;... 4.25MB
Title Monte Carlo evaluation of the dosimetric uncertainty in matched 6 MV Elekta and Varian linear accelerators
Author(s) Lye, J
Year 2012
Abstract To facilitate a national improvement in therapeutic dosimetry, ARPANSA has installed a clinical linear accelerator (linac) to provide chamber calibrations at the user beam quality. The ARPANSA reference linac is an Elekta Synergy platform linac, which can be tuned to mimic the beam qualities available in Australian clinics, such as those produced by the Varian linac platforms.

Monte Carlo models of a Varian and matched Elekta accelerator have been developed and commissioned. They accurately predict the measured percentage depth dose and profiles, but show significantly different energy spectra, resulting mainly from differences in target thickness between the two accelerators. The critical question is if the spectral difference between the Varian and the matched Elekta will cause any significant change in a chamber calibration factor. This work calculates the error introduced when using a calibration factor obtained from this Elekta Synergy Platform linac on a Varian high energy platform beam at 6 MV.

To determine the answer a two stage modelling approach was applied. Firstly the graphite calorimeter that is the Australian primary standard of absorbed dose was modelled. Using the commissioned linac models, the calorimetry measurements of absorbed dose to graphite can be converted to absorbed dose to water in a direct Monte Carlo conversion. Using the direct MC conversion removes from the calorimeter response any dependence on the source spectrum. The ionisation chamber calibration factor is then only dependant on the chamber response. The direct MC conversion without a transfer chamber in graphite is also a novel method of realising absorbed dose to water from graphite calorimetry measurements.

The second stage was to model a secondary standard chamber type NE2561. The modelling of the energy correction factor kQ of a secondary standard NE2561 chamber shows a difference of 0.4 % between the Varian and Varian-matched Elekta beams. Although small, this is a significant discrepancy for primary standard calibrations. Ionisation chambers calibrated on the ARPANSA 6 MV beam may be in error by 0.4% when used on a Varian 6 MV beam. Similar variations may occur with other manufacturers. The work has also investigated the design of a custom flattening filter to precisely match the energy spectrum of the Varian beam on the Elekta platform.

The Monte-Carlo investigation of the kQ of the secondary standard NE2561 chamber was extended to consider the response with the new flattening filter free (FFF) modality. The kQ from a modelled Varian FFF beam showed a difference of 0.8 % compared to the 6 MV Elekta with the same beam quality index, TPR20,10, (Tissue phantom ratio in water at depths of 20 and 10 cm), and the Elekta FFF beam showed a difference of 0.6 %.

The magnitude of the discrepancy between the Varian FFF beam and the ARPANSA Elekta beam demonstates that TPR20,10 is not sufficient as the sole beam quality specifier to derive calibration factors for flattening filter free beams, and even with standard clinical beams the use of TPR20,10 alone can introduce errors of 0.4 %.
Degree Masters by Research
Institution RMIT University
School, Department or Centre Applied Sciences
Keyword(s) Monte Carlo modelling
linac
primary standard of absorbed dose
kQ
flattening filter free
Versions
Version Filter Type
Access Statistics: 381 Abstract Views, 1073 File Downloads  -  Detailed Statistics
Created: Mon, 29 Apr 2013, 14:04:33 EST by Brett Fenton
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us