Studying the potential of point detectors in time-resolved dose verification of dynamic radiotherapy

• Time-resolved dose measurements during highly modulated radiotherapy are presented.
• Plastic scintillators of different size are compared with ionization chambers.
• Detector response in dynamic modulated fields are dominated by volume averaging.
• Differences between detectors are averaged out with accumulated dose.
• Measured dose to a small degree depend on detector temporal resolution.

Modern megavoltage x-ray radiotherapy with high spatial and temporal dose gradients puts high demands on the entire delivery system, including not just the linear accelerator and the multi-leaf collimator, but also algorithms used for optimization and dose calculations, and detectors used for quality assurance and dose verification. In this context, traceable in-phantom dosimetry using a well-characterized point detector is often an important supplement to 2D-based quality assurance methods based on radiochromic film or detector arrays. In this study, an in-house developed dosimetry system based on fiber-coupled plastic scintillator detectors was evaluated and compared with a Farmer-type ionization chamber and a small-volume ionization chamber. An important feature of scintillator detectors is that the sensitive volume of the detector can easily be scaled, and five scintillator detectors of different scintillator length were thus employed to quantify volume averaging effects by direct measurement. The dosimetric evaluation comprised several complex-shape static fields as well as simplified dynamic deliveries using RapidArc, a volumetric-modulated arc therapy modality often used at the participating clinic. The static field experiments showed that the smallest scintillator detectors were in the best agreement with dose calculations, while needing the smallest volume averaging corrections. Concerning total dose measured during RapidArc, all detectors agreed with dose calculations within 1.1 ± 0.7% when positioned in regions of high homogenous dose. Larger differences were observed for high dose gradient and organ at risk locations, were differences between measured and calculated dose were as large as 8.0 ± 5.5%. The smallest differences were generally seen for the small-volume ionization chamber and the smallest scintillators. The time-resolved RapidArc dose profiles revealed volume-dependent discrepancies between scintillator and ionization chamber response, which confirmed that correction factors for ionization chambers in high temporal and spatial dose gradients are dominated by the volume averaging effect. The unique scaling of the scintillator volumes indicated how such time-dependent volume averaging corrections could be quantified. The time-resolved measurements further supported the claim that small-volume water equivalent detectors are most likely to accurately detect changes in dose delivery, although exact positioning of detectors remains critical.