Three dimensional film dosimetry of photon beam in small field sizes and beyond the heterogeneous regions using a GAFchromic films array

Abstract

Introduction: Inaccurate dose prediction by treatment planning system (TPS) beyond the heterogeneous regions has been a great challenge in radiotherapy, particularly when small field sizes are used. Better understanding of this inaccuracy that depends on dosimetry methods is very crucial. Recently, the film stack dosimeter has been purposed for 3D comprehensive dosimetry. This study proposes an approach to investigate the accuracy of TPS for predicting dose distribution in small field sizes in the presence of nasopharynx heterogeneities by 3D stack film. Material & methods: The dose distribution and dose profile of rectangular heterogeneous nasopharynx (RHN) phantom with two air cavities and bone equivalent were measured with a 3D stack film dosimeter containing 9 GAFchromic-EBT2 films positioned beyond the heterogeneity regions. The film was validated using MCNPX-Monte Carlo (MC) simulation. The evaluation of TPS inaccuracy in small field sizes (3 x 3, 4 x 4, 5 x 5 cm(2)) was determined by comparing reconstructed 3D dose distribution of stack film dosimeter with TPS based on full scatter convolution (FSC) using in-house Matlab code. Results: The film measurements had a good agreement with the MC calculation for small fields in RHN phantom. The results showed a large discrepancy between stack film measurements and TPS calculations in the volumes enclosed high isodoses, so that the biggest difference occurred in 3 x 3 cm(2) field size (relative mean difference = 1.72 +/- 0.24, 1.5 +/- 0.16 and P-value = 0.002, 0.001 for V95 and V90 respectively) and it decreased in 5 x 5 cm(2) field size (relative mean difference = 0.46 +/- 0.05, 0.35 +/- 0.04 and P-value = 0.001, 0.002 for V95 and V90 respectively). Conclusion: The results suggested that 3D stack film dosimetry can be successfully used as a reliable dosimeter for QA procedure of heterogeneous region in small field sizes. Comparison of the measured and calculated dose volumes demonstrated that the TPS based FSC algorithm has a notable inaccuracy in estimation of dose distribution beyond the nasopharynx heterogeneities. Therefore: it seems heterogeneity corrections should be considered for nasopharynx treatment.