Landau distribution-based regularized algorithm for reconstruction of electron beam energy spectrum using depth dose distributions in targeted materials

Abstract

Assuming the beam spectrum at the accelerator output follows a Gaussian distribution and its attenuation in air is described by a Landau distribution, the ill-posed problem of electron energy spectrum reconstruction using depth-dose distribution was reduced to finding the electron spectrum at the phantom surface as a convolution of the Landau and Gaussian distributions. The algorithm proposed in the study uses the experimentally measured depth dose distributions generated by Varian TrueBeam medical accelerator operating at 6 MeV and 9 MeV modes in water-equivalint and aluminium phantoms and reconstructs the electron spectra in targeted materials. The algorithm uses reference depth dose distributions from monoenergetic electrons calculated using Geant4 toolkit. It was found that the spectra reconstructed using experimental depth dose distributions in solid water and aluminium differ by less than 5%, and reconstructed depth dose distributions corresponding to the reconstructed spectra deviate from experimentally measured ones by no more than 5%.