An evaluation of the stability of image‐quality parameters of Varian on‐board imaging (OBI) and EPID imaging systems

Quality assurance (QA) of the image quality for image-guided localization systems is crucial to ensure accurate visualization and localization of regions of interest within the patient. In this study, the temporal stability of selected image parameters was assessed and evaluated for kV CBCT mode, planar radiographic kV, and MV modes. The motivation of the study was to better characterize the temporal variability in specific image-quality parameters. The CATPHAN, QckV-1, and QC-3 phantoms were used to evaluate the image-quality parameters of the imaging systems on a Varian Novalis Tx linear accelerator. The planar radiographic images were analyzed in PIPSpro with high-contrast spatial resolution (f30, f40,f50 lp/mm) being recorded. For OBI kV CBCT, high-quality head full-fan acquisition and pelvis half-fan acquisition modes were evaluated for uniformity, noise, spatial resolution, HU constancy, and geometric distortion. Dose and X-ray energy for the OBI were recorded using the Unfors RaySafe Xi system with the R/F High Detector for kV planar radiographic and the CT detector for kV CBCT. Dose for the MV EPID was recorded using a PTW975 Semiflex ion chamber, PTW UNIDOS electrometer, and CNMC Plastic Water. For each image-quality parameter, values were normalized to the mean, and the normalized standard deviations were recorded to evaluate the parameter's temporal variability. For planar radiographic modes, the normalized standard deviations of the spatial resolution (f30, f40, & f50) were 0.015, 0.008, 0.004 lp/mm and 0.006, 0.009, 0.018 lp/mm for the kV and MV, respectively. The normalized standard deviation of dose for kV and MV were 0.010 mGy and 0.005mGy, respectively. The standard deviations for full- and half-fan kV CBCT modes were averaged together. The following normalized standard deviations for each kV CBCT parameter were: 0.075 HU (uniformity), 0.071 HU (noise), 0.006mm (AP-geometric distortion), 0.005 mm (LAT-geometric distortion), 0.058mm (slice thickness), 0.124 (f50), 0.031 (HU constancy - Lung), 0.063 (HU constancy- Water), 0.020 (HU constancy - Bone), 0.006 mGy (Dose - Center), 0.004 mGy (Dose -Periphery). Using control chart analysis, institutional QA tolerances were reported as warning and action thresholds based on 1σ and 2σ thresholds. A study was performed to characterize the stability of image-quality parameters recommended by AAPM Task Group-142 for the Varian OBI and EPID imaging systems. Both imaging systems show consistent imaging and dosimetric properties over the evaluated time frame.