Session Item

Bragg Peak range verification is an important part of routine quality assurance for proton beam therapy centres. We evaluated the suitability of a commercial scintillator-based device for routine range verification checks on a Varian Probeam system (Varian Inc, USA). Commercial scintillator devices are fast, compact and portable relative to competing technologies like multi-layer ionisation chambers. Signal processing is arguably more complex for scintillators, however. These devices must contend with factors such as: optical distortion, quenching, light reflection/refraction within the instrument, and water equivalent thickness variations.

A commercial scintillation device, the Logos Ranger-300 (Logos Systems Int'l, USA), was used to measure Bragg Peak ranges for 19 energies between 70-245 MeV. Bragg Peak range was defined as the 80% distal drop off. In total 16 measurement sessions were conducted across four gantries. The results from these sessions were compared against NIST reference values, water tank (WT) data used to define our beam model and 16 independent measurement sessions using an IBA Giraffe multi-layer ionisation chamber (IBA Group, Belgium). Ranger and Giraffe Accuracies were assessed by their respective standard deviations across sessions. Precision was assessed by comparing our measurements to NIST and WT reference datasets, calculating mean range differences and standard deviations across all sessions. Range means and 95% confidence intervals were also calculated per energy. Range changes were measured by placing solid water in the beam path and the effects of setup error assessed by deliberately misaligning the Ranger to the beam-axis in additional measurement sessions. In-house software was used to analyse Ranger data, correcting for some non-linear distortions and scintillator water equivalent thickness. It is available for research purposes: https://github.com/UCLHp/RangerQA.

Ranger accuracy was comparable to Giraffe, both exhibiting a 0.15 mm standard deviation (SD) over all measurements. Figure 1a shows measurement differences from reference data. Ranger accuracy was consistent across energies, whilst Giraffe was inversely proportional (Fig 1b). Ranger precision was comparable to Giraffe across energies, both giving mean (SD) absolute differences ≤0.25 (0.22) mm when compared to WT and NIST reference values. Compared to reference data, both devices had a 95% CI measurement uncertainty within +/-0.25 mm for all energies. The largest mean disparity for any given energy was within 0.7 mm for WT and 0.8 mm for NIST on both devices. Ranger was sensitive to 1 mm incremental increases of solid water thickness in the beam path (Fig 2), but robust to misalignments of >1mm and >1 degree.

Commercial scintillator range verification devices are capable of monitoring depth changes well within the ±1 mm threshold set out in AAPM TG224. They are robust to setup errors and offer a compelling alternative to multi-layer ionisation chambers.