Session Item

Cherenkov emission carries the potential of direct, perturbation-free, in water dose measurements. However, until now such measurements suffered from large (up to 60%) uncertainties because of the intrinsic anisotropy of Cherenkov emission. Cherenkov radiation is emitted along a cone whose angle is determined by the charged particles energy and direction which vary with tissue attenuation making corrections non-trivial. This work investigates the use of polarization imaging  to precisely measure and correct electron beam dose distributions. 

Cherenkov emission produced in a 15 x 15 x 15 cm3 water tank, from 6 MeV and 18 MeV electron beams, is measured by a CCD camera (414EX; Atik Cameras, Norwich, United Kingdom) coupled to a rotating polarizer. Images are acquired from four polarization axis (0°, 45°, 90°, 135°), as presented on figure 1. The polarized component of the signal is extracted using Malus law. This polarized portion should be proportional to the dose, given a correction factor that can be defined by Cherenkov polar and azimuthal angular distributions. Hence, Cherenkov emission was simulated using the Geant4 toolkit (v4.10.04). Monte Carlo angular distributions were then used to correct polarized measurements. The resulting images consist of 2D projections as the signal is summed along the camera’s optical axis. Projected percent depth dose (PPDD) and profiles were extracted for each measurement and compared to equivalent dosimetric film measurements.

Cherenkov emission generated from electron beams exhibited a strong polar dependence, peaked at 41°, with a distribution that varied with depth. This resulted in large dose discrepancies (up to 60%) using the raw Cherenkov emission for dose measurements. Overall, 42 ± 4% (6 MeV) and 47 ± 3% of the signal was found to be linearly polarized. After applying a polarization correction,  PPDD presented differences (mean ± standard deviation) with film measurements of −1 ± 2% (6 MeV) and 0.9 ± 2% (18 MeV). For projected profiles, differences of 0.9 ± 0.6% (6 MeV) and 0.8 ± 0.6% (18 MeV) between films and polarized corrected signals are obtained on the beam central axis. Figure 2 summarizes the dose measurement results.

The results suggest that accurate, perturbation-free in water dose measurement can be achieved using Cherenkov radiation combined with polarization imaging.  Since the correction function is based on angular distributions, which are not directly related to the dose, the method is expected to be less prone to redundancy errors than previously proposed Monte Carlo based correction methods.