Routine kV-CBCT quality assurance in IGRT: Workflow-based comparison of QUART/MaximQA vs Catphan/ARTISCAN

The study aimed to compare image-quality metrics measured by two kV-CBCT QA systems-the QUART phantom with MaximQA and the Catphan phantom with ARTISCAN-using standard clinical kV-CBCT imaging protocols in a Varian TrueBeam IGRT setting, in line with AAPM TG-142/TG-179 QA recommendations.

QA scans of the QUART and Catphan phantoms were acquired for image gently, head, thorax, pelvis and pelvis large imaging protocols. Key image quality metrics (noise, signal-to-noise ratio [SNR], uniformity, geometric distortion, Hounsfield unit [HU] constancy, contrast-to-noise ratio [CNR]) were measured using MaximQA and ARTISCAN software. Descriptive statistics and paired t-tests (α = 0.01) evaluated differences between systems; TG-142/TG-179 tolerances provided context.

The QUART/MaximQA system consistently produced significantly lower measured image noise and higher measured SNR than Catphan/ARTISCAN across most protocols (e.g., Pelvis Large: SNR 495.1 ± 40.7 vs 283.6 ± 31.5, p < 0.01). Geometric accuracy was excellent in both systems (measured distances ∼99.6%-100% of nominal, no significant differences). QUART phantom HU values were closer to nominal for inserts; however, the Catphan Solid Water insert was interpreted in terms of material-related HU behavior and stability rather than poorer HU accuracy (e.g., water-equivalent insert: QUART ≈ 0 HU vs +60 for Catphan/Solid Water, p < 0.01). Uniformity differences were small (< 5 HU) and often not significant. CNR and low-contrast detectability were reported as complementary QA indicators rather than directly equivalent metrics. Overall, both systems yielded image quality metrics within TG-142/TG-179 baselines.

Both QUART/MaximQA and Catphan/ARTISCAN provide consistent kV-CBCT QA metrics, but due to differences in phantom design, ROI definition, analysis software and setup conditions, their absolute values should not be considered directly interchangeable. The choice of phantom/software may influence absolute QA values. We recommend using institution-specific baselines per TG-142/TG-179 and cite emerging literature on automated kV-CBCT QA.