ESTRO 38 Abstract book

S147 ESTRO 38

Calibration Calibration

0.7% 0.1%

0.7% 0.1%

factor

Calibration stability

Conversion kC kC assignment

0.50% 0.40%

0.53% 0.84%

Influence quantities Linac stability

0.05% 1.1%

0.05% 1.3%

COMBINED (k=1)

Conclusion The excellent agreement between MC and experiment motivate CE-based dosimeter design efforts. Based on these results, we recommend a large-aperture detector and downstream PDC shift for beam-axis CE-based electron beam dosimetry in water. This is an attractive alternative to current methods as it avoids perturbations, can be extended to 3D via tomography or optical sectioning, and is especially promising for small fields due to the high resolution achievable with optical methods. OC-0294 Separating initial and general recombination in reference dosimetry of proton pencil beam scanning J.B. Christensen 1 , E. Almhagen 2,3 , L. Stolarczyk 2,4 , M. Liszka 4 , A. Vestergaard 5 1 Technical University of Denmark, Center for Nuclear Technologies, Roskilde, Denmark ; 2 The Skandion Clinic, The Skandion Clinic, Uppsala, Sweden ; 3 Uppsala University, Medical Radiation Sciences- Department of Immunology- Genetics and Pathology, Uppsala, Sweden ; 4 Institute of Nuclear Physics- Polish Academy of Sciences in Krakow, The Bronowice Cyclotron Centre, Krakow, Poland; 5 Aarhus University Hospital, Danish Centre for Particle Therapy, Aarhus, Denmark Purpose or Objective Gas-filled ionization chambers are the detectors of choice for calibrating ion beams at proton therapy facilities. The charge liberated in the ionization chamber is collected via an applied voltage and related to the absorbed dose. However, ion pairs of opposite charge may recombine and lead to an underestimation of the dose. The recombination events cannot be corrected with Monte Carlo methods as the recombination cross sections contain too large uncertainties. As a result, recombination losses are traditionally corrected with methods dating anywhere from decades to a century back. The main objective is to obtain a more precise determination of the recombination correction by separation of initial and general recombination for use in reference dosimetry. Material and Methods A Roos type parallel-plate ionization chamber was irradiated with a continuous scanned proton beam at 70, 150, and 226 MeV and 4 dose rates at in a water phantom. The liberated charge was for each configuration collected at 5 polarization voltages between 50 and 200 V. A detailed Monte Carlo model of the beam line enables an accurate calculation of both the particle spectrum and a conversion from delivered MU/minute to dose rate as the beam scans over the ionization chamber. The charge collection at several dose rates in turn permits a calculation of the recombination as a function of dose rate using the Boutillon theory [1]. The results are compared to the Jaffé theory [2] for initial recombination, the two- voltage-method (TVM) suggested in TRS-398 [3], and recombination corrections using extrapolation methods. Results The experimentally determined recombination correction factors calculated with the TVM and extrapolation methods agree within 1.5 %. The TVM data are in the figure plotted as a function of the calculated dose rate for a collecting voltage of 200 V. The dotted line represents the Boutillon theory for initial and general recombination

Fig. 1: Cherenkov detection setup. Not to scale.

Results MC and measured relative kC factors agree to within 1% for percent-depth CE (PDC)>50% (Fig. 2). At other depths, deviations are in accordance with approximations. Simulated electron beam quality specifier R50 is estimated from the depth of 50% CE C50 to within 0.1 mm and 0.4 mm (maximum) with large and small apertures respectively. The fit performance on measured R50 supports this finding. The kC value is uniquely specified by R50 at a given depth with preliminary dosimetric uncertainty estimate of the order of 1% (Table 1). A PDC downstream shift by the R50-C50 difference is derived from theory and found to reduce the kC depth dependence tenfold to 2.5%.

Fig. 2: (a) Simulated (Sim) and experimental (Exp) percent-depth dose, PDD, percent-depth Cherenkov emission (CE), PDC, at 90 degrees to beam, and (b) PDD/PDC (i.e., normalized CE-to-dose conversion, kC) in water from 20 MeV TrueBeam electrons. The error bars define an interval estimated to have 95% level of confidence. Table 1: Preliminary best-case uncertainty budget of Cherenkov emission-based measurement of absorbed dose to water in electron beams at optimal depth dref = 0.6R50 - 0.1 [cm] under reference conditions.

4π detection 90º±5º detection

Measurement M_raw SSD Position Field sizeTemperature

0.3% 0.10% 0.33% 0.10% 0.06%

0.3% 0.10% 0.33% 0.10% 0.06%