Jiang, Steve B

Abstract Background The RefleXion X1 is a novel PET/CT‐guided ring gantry linac that introduces unique geometry and workflow compared to conventional systems. Existing QA guidelines (AAPM TG‐142, 148, 306) cover standard linacs and tomotherapy but do not address this platform's needs. Therefore, an automated daily QA solution is required to ensure accurate output, imaging guidance, and couch positioning, which motivated the development and evaluation of the workflow presented in this study. Purpose To develop and evaluate an automated, efficient, and comprehensive daily quality assurance (QA) workflow for a novel PET/CT‐guided ring gantry linear accelerator (RefleXion X1), incorporating tests for dosimetry, imaging guidance, and couch positioning accuracy. Methods An IMRT QA plan was designed following the CT scan of the TomoDose 2D diode detector array (Sun Nuclear, Melbourne, FL) within the RefleXion treatment planning system (TPS). Several spherical targets of varying sizes were contoured to conform the dose distribution to these targets and evaluate beam output, laser localization, profile constancy, and CT‐MV isocenter coincidence. Linear models correlating spatial location of dose peak with intentional shifts in different directions were developed and integrated into an in‐house analysis software, iQA, written in C#. The software was validated for robustness and successfully implemented in clinical practice. Results Validation with 42 intentional shifts showed submillimeter accuracy, with mean deviations of 0.2 mm (X), 0.1 mm (Y), and 0.79 mm (Z), and R 2 values ≥ 0.98. Over 1 year, daily QA data showed stable output constancy (1.006 ± 0.016), left/right target dose constancy (1.007 ± 0.005 and 0.996 ± 0.004), and consistent CT‐MV isocenter offsets (–0.335 ± 0.23 mm in X, 0.19 ± 0.19 mm in Y, and –0.544 ± 0.36 mm in Z). The software also detected a gradual output drift in between, prompting a timely linac adjustment. Conclusion This fully automated, single‐delivery QA workflow enables efficient and robust evaluation of multiple IGRT performance parameters. It improves standardization, reduces human error, and reliably monitors long‐term machine stability. The tools and methodology may be adopted by other clinics seeking a practical and automated daily QA solution. Future developments will focus on incorporating PET QA and rotational couch validations to further enhance the system's comprehensiveness and adaptability.

AbstractBackgroundThe growing adoption of magnetic resonance imaging (MRI)‐guided radiation therapy (RT) platforms and a focus on MRI‐only RT workflows have brought the technical challenge of synthetic computed tomography (sCT) reconstruction to the forefront. Unpaired‐data deep learning‐based approaches to the problem offer the attractive characteristic of not requiring paired training data, but the gap between paired‐ and unpaired‐data results can be limiting.PurposeWe present two distinct approaches aimed at improving unpaired‐data sCT reconstruction results: a cascade ensemble that combines multiple models and a personalized training strategy originally designed for the paired‐data setting.MethodsComparisons are made between the following models: (1) the paired‐data fully convolutional DenseNet (FCDN), (2) the FCDN with the Intentional Deep Overfit Learning (IDOL) personalized training strategy, (3) the unpaired‐data CycleGAN, (4) the CycleGAN with the IDOL training strategy, and (5) the CycleGAN as an intermediate model in a cascade ensemble approach. Evaluation of the various models over 25 total patients is carried out using a five‐fold cross‐validation scheme, with the patient‐specific IDOL models being trained for the five patients of fold 3, chosen at random.ResultsIn both the paired‐ and unpaired‐data settings, adopting the IDOL training strategy led to improvements in the mean absolute error (MAE) between true CT images and sCT outputs within the body contour (mean improvement, paired‐ and unpaired‐data approaches, respectively: 38%, 9%) and in regions of bone (52%, 5%), the peak signal‐to‐noise ratio (PSNR; 15%, 7%), and the structural similarity index (SSIM; 6%, <1%). The ensemble approach offered additional benefits over the IDOL approach in all three metrics (mean improvement over unpaired‐data approach in fold 3; MAE: 20%; bone MAE: 16%; PSNR: 10%; SSIM: 2%), and differences in body MAE between the ensemble approach and the paired‐data approach are statistically insignificant.ConclusionsWe have demonstrated that both a cascade ensemble approach and a personalized training strategy designed initially for the paired‐data setting offer significant improvements in image quality metrics for the unpaired‐data sCT reconstruction task. Closing the gap between paired‐ and unpaired‐data approaches is a step toward fully enabling these powerful and attractive unpaired‐data frameworks.