Cherenkov emission‐based external radiotherapy dosimetry: I. Formalism and feasibility

Purpose Cherenkov emission (CE)‐based external beam dosimetry is envisioned to involve the detection of CE directly in water with placement of a high‐resolution detector out of the field, avoiding perturbations encountered with traditional dosimeters. In this work, we lay out the groundwork for its implementation in the clinic and motivate CE‐based dosimeter design efforts. To that end, we examine a formalism for broad‐beam in‐water CE‐based dosimetry of external radiotherapy beams, design and test a Monte Carlo (MC) simulation framework for the calculation of CE‐to‐dose conversion factors used by the formalism, and demonstrate the experimental feasibility of this method. Methods The formalism is conceptually analogous to ionization‐based dosimetry and employs CE‐to‐dose conversion factors, kCθ±δθ, including only and all CE generated within polar angles θ ± δθ on beam axis. The EGSnrc user code SPRRZnrc is modified to calculate kCθ±δθ, as well as CE spectral and angular distributions. The modified code is tested with monoenergetic parallel electrons on a thin water slab. Detector configurations are examined for broad 6–22 MeV electron beams from a BEAMnrc TrueBeam model, with a focus on θ±δθ=90∘±90∘ (4π detection), 90∘±5∘, and 42∘±5∘ (θ=42∘ is the CE angle of relativistic electrons in water). We perform a relative experimental validation at 90∘ with electron beams, using a simple detector design with spherical optics and geometrical optics approximation of the sensitive volume, which spans the water tank. Due to transient charged particle equilibrium, broad photon beams are generally less sensitive to beam quality, depth, and angle. Results For 0.1–50 MeV electrons on a thin water slab, the code outputs CE photon spectral density per unit mass (calculated from dose and kCθ±δθ) and angle in agreement with theory within ±0.03% and ±0.01∘, respectively, corresponding to the output precision. The 42∘ configuration was found impractical due to detection considerations. Detection at 90∘±δθ for small δθ exhibited beam quality dependence of the same order as well as strong superficial depth dependence. A 4π configuration ameliorates these effects. A more practical approach may employ a large numerical aperture. In comparing with literature, we find that these effects are less pronounced for broad photon beams in water, as expected. Measured relative kC90∘±δθ at small δθ were within 1% of simulated factors (relative to their local average) for percent‐depth CE (PDC