Plastic-scintillator based dosimetry for minibeam radiotherapy

Minibeam radiotherapy (MBRT) is a dose delivery technique that exploits the alternation of high-dose ("peaks") and low-dose ("valleys") regions, with the potential goal of achieving a higher sparing of healthy tissues without compromising tumor control compared to conventional radiotherapy. The involved beamlets are millimetric, and single temporal fractions are used. Ultra High Dose Rate (UHDR) radiotherapy is another modality that offers the potential to achieve these outcomes by eliciting the so-called FLASH effect, a peculiar tissue response generated by ultra-fast beam delivery at average dose rates thousands of times larger than conventional radiotherapy. Both irradiation methods present significant challenges from a dosimetric perspective: UHDR beams often induce a non-linear response in dosimeters, due to the high deposited dose in a brief time interval, and MBRT requires fine spatial resolution to resolve the beamlet structures. Plastic-scintillator based dosimeters may be suitable candidates for both irradiation modalities, given their potentially high spatial resolution and linearity with dose and dose rate, besides being almost tissue-equivalent and more cost-effective than other detector technologies. The main objective of this thesis was to assess the dosimetric performance of a plastic scintillator based detector for low-energy electron FLASH beam and minibeam radiotherapy. The developed dosimeter consisted of a 0.5-mm thick sheet of EJ212 plastic scintillator placed in a custom PVC holder and irradiated with 9-MeV electrons, generated by the Triode-Gun equipped ElectronFlash LINAC, available at the Centro Pisano for FLASH Radiotherapy (CPFR). A 45◦ inclined mirror was used to reflect the scintillation light emitted by the plastic sheet toward a scientific CCD camera coupled with an optical lens. The performance of the dosimeter was evaluated by assessing the dose-response linearity, the capability to measure the percentage depth dose in solid water, the spectrum of the plastic scintillator emission, the ability to measure minibeam dosimetric parameters, and the spatial resolution of the setup. The linearity was tested by varying the beam Dose Per Pulse (DPP). The percentage depth dose curve was measured at a fixed output dose by adding layers of solid water with increasing thickness before the scintillator. In this case, the images were acquired both unfiltered and with three optical filters (centered at 405 nm, 457 nm, and 532 nm) to assess the Cerenkov radiation contribution and correct the PDD to obtain a better agreement with the reference curve (measured with a calibrated FLASH diamond detector). The EJ212 spectral emission was analyzed with a spectrograph and fiber-optic system. The electron beam was placed at various depths in solid water, and a sufficiently low-energy X-ray beam was used to measure the scintillation spectrum alone without Cerenkov emission. The capability to image spatially-fractionated dose patterns was investigated, with different minibeam collimators, varying the DPP, and the depths in solid water. Three parameters were evaluated: the peak-valley dose ratio (PVDR), the mean center-to-center (CTC) distance of adjacent peaks, and the peak full-width-half-maximum (FWHM). Finally, the spatial resolution of the scintillator and imaging system was evaluated using an X-ray source and images of an edge. The contribution of the imaging system alone (with no scintillator) was estimated using optical patterns. Linearity was found to be verified up to 12 Gy per pulse, corresponding to 3 MGy/s of intra-pulse dose rate. The PDD was correctly measured with the dosimeter. The Cerenkov correction enabled a better agreement with the reference PDD, reducing the residuals from 6% to 3% vs. the reference. All the spectra were mainly coherent with the tabulated EJ212 emission (peak wavelength at 423 nm). A mismatch with the tabulated spectrum was found around 405 nm, presumably due to thickness-dependent self-absorption effects. Moreover, the Cerenkov light contribution could be observed above 500 nm in the spectrum obtained with electrons. The spatial features of fractionated beams (CTC and FWHM) were correctly reconstructed within ∼ 6% of the reference values. For the PVDR values, since no reference simultaneous measurements were acquired, the only comparison could be made with a recently published paper, which seems to suggest slightly larger values than the ones measured in this work. This difference could be partially due to current limits in the spatial resolution of the imaging system and to slight differences in the data analysis. However, the main contributor is presumably a technical intervention on the LINAC between the two measurements, which affected the beam energy spectrum and the delivered dose. The spatial resolution was found to be in the tenths of a millimeter range. The results suggest that this setup based on plastic scintillator sheets is appropriate for low-energy electron FLASH and MBRT dosimetry applications. Future improvements will be related to improving the spatial resolution, validating the measurements of dose spatial fractionation with a reference detector, and better correcting the Cerenkov contribution.