Room: AAPM ePoster Library
Purpose: To evaluate the ability of simulated single energy (SE) and dual energy (DE) projections, derived from CT images, to predict the accuracy of markerless tumor tracking (MTT).
Methods: A motion phantom, consisting of a torso with embedded ribs/spine, along with a cavity having lung-equivalent density was used in this study. Three targets with diameters of 5, 10 and 15mm were individually placed in the lung-equivalent cavity. For each target, three different static positions were used representing the two extremes and the mid-point of respiratory motion. Using each combination of target size and location, planning CT scans were obtained. Simulated projections (60 and 120kVp) were created from each of these CT scans. DE projections were produced through weighted logarithmic subtraction. Separately, cone beam computed tomography (CBCT) acquisitions were obtained using SE and fast-kV switching DE imaging with the on-board imager (OBI) of a commercial linear accelerator. A template-based matching algorithm was used to track target motion on both simulated and actual SE/DE images. Tumor tracking coordinates were evaluated against ground truth (for each target) using the receiver operating characteristics (ROC) and root-mean-squared error (RMSE – for 95% specificity) on the combined results for the three tumor positions.
Results: For the 5mm target, the ROC area-under-the-curve (AUC) were 0.902 and 0.886 while RMSE were 1.69mm and 1.55mm, for DE actual and simulated images, respectively. For SE images, the AUC were 0.797 vs. 0.776 (actual vs. simulated), and RMSE were 2.11mm vs. 2.15mm (actual vs. simulated). For all targets considered, the correlation coefficients between actual and simulated images were 0.96 and 0.95 for AUC and RMSE, respectively.
Conclusions: Simulated projections, derived from CT, may be used to predict the accuracy of MTT at the time of planning. This may allow for the selection of optimal imaging modality (SE vs. DE) for MTT.
Funding Support, Disclosures, and Conflict of Interest: Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under Award Number R01CA207483. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.