Quantitative image quality evaluation for kV cone-beam CT-based IGRT

The objective of this study is to quantitatively evaluate the image quality of a kV cone-beam CT-based IGRT system (Elekta, XVI) using two commercial CT image quality phantoms, Catphan-600 and CIRS-062QA. Both phantoms consist of similar image quality test modules (uniformity, CT linearity and spatial resolution) but each phantom has different diameter and test pattern design. Each test module was imaged separately using an optimised cone-beam CT imaging parameter. The quality metrics of the reconstructed images were analysed using algorithms developed with MatLab. The image uniformity and the spatial resolution measured with Catphan were of 4% and 40% greater respectively, compared to those measured with CIRS phantom. The differences were due to the beam scattering and hardening originated from the CIRS phantom holder. The contrast-to-noise ratio (CNR) values measured with CIRS phantom were at least 2% higher than that of Catphan. The diameter of CIRS phantom is smaller and resulted in lower beam attenuation. The quantitative image quality assessment algorithms developed for both phantoms provided a phantom-specific set of reference values for a cone-beam CT imaging system as recommended by AAPM TG-142. Further investigation will be performed to resolve beam hardening issue arising from the CIRS phantom holder.