Optimization of the iterative deconvolution correction method applied to ionization chamber response for small field dosimetry measurements

Abstract

This study aims to optimize the iterative deconvolution correction of air-filled ionization chamber measurements with limited spatial resolution for photon beam dosimetry. The ionization chamber volume effect could be explained by the inhomogeneity of the absorbed dose in its sensitive volume, leading to a volume averaging effect acting as a convolution kernel K(x). Therefore, the “true” dose profile Pt(x) can be deduced by deconvolving measured 1D profiles Pm(x). The Semiflex 3D ionization chamber type PTW 31021 was studied for beams with flattening filter (WFF), while the SNC125c ionization chamber was investigated for flattening filter free (FFF) beams. The deconvolution distributions were compared for the first ionization chamber with reference dose profiles calculated by Monte Carlo simulation and for the second ionization chamber with reference measurements obtained using a diode EDGE detector, suitable for small field applications. The convolution kernel was assumed to be a sum of normalized Gaussian and Lorentz distributions parametrized by the pair (σicλic ). Good results were obtained with respect to the γ index 2.0%/0.5 mm criterion for the field sizes 6 × 6 mm2, 10 × 10 mm2 and 20 × 20 mm2. The highest agreement between reference and corrected measurement data was obtained by using the Gauss-Lorentz distribution parameters (1.92 mm, 0.80 mm) and (1.90 mm, 0.90 mm) for the ionization chambers Semiflex 3D type PTW 31021 and SNC125c, respectively.