Evaluating the suitability of diagnostic CT for oART planning: Impact of kV and phantom selection

This study investigates the use of diagnostic CT (dCT) images acquired at different x-ray tube voltages for simulation-free online adaptive radiotherapy (oART). Specifically, we evaluated the dosimetric impact of CT peak voltage (kV_p) and phantom selection on HU-to-density calibration for IMRT/VMAT planning, using two dose calculation algorithms: Acuros XB (AXB) and Anisotropic Analytical Algorithm (AAA).

HU-to-relative electron density (RED) and HU-to-mass density (MD) calibration curves were generated using CIRS 062 M and CatPhan 504 phantoms scanned at 80-140 kV_p on a Siemens go.Sim CT. 20 palliative 6MV-FFF VMAT/IMRT plans were recalculated in Eclipse v16 using AXB and AAA. Two dosimetric comparisons were conducted: (1) evaluating the effect of kV_p using kV_p-specific calibrations (kV_p-120 as reference) for each phantom, and (2) contrasting phantom types across kV_p levels. Evaluation included DVH metrics (PTV-V_95%, V_105%, D_median-OAR mean doses, body D_max, conformity indices), and gamma analysis (2%/2 mm, 1%/1 mm). Statistical significance was assessed using Wilcoxon and two-way ANOVA (α = 0.05).

kV_p-induced HU variations in CatPhan were minimal (unlike CIRS), resulting in negligible dosimetric impact for both phantoms and algorithms (<1%). AXB showed slightly greater robustness. CatPhan-based calibrations consistently underestimated DVH parameters versus CIRS, especially in high-density tissues. Median gamma passing rates exceeded 99 % (2%/2 mm) and 98 % (1%/1 mm) for kV_p analysis, but dropped to 88-95 % (1%/1 mm) when comparing phantom types.

dCT at different kV_p settings is dosimetrically suitable for IMRT/VMAT planning. These findings reinforce best practices in CT calibration, essential for implementing promising resource-efficient workflows such as simulation-free oART.