Room: 225BCD
Purpose: We assess Hounsfield Unit (HU) accuracy and noise as well as CNR for low dose megavoltage cone-beam acquisition (MVCBCT) using a novel, high DQE multi-layer imager (MLI).
Methods: The MLI prototype is composed of four-layers, each consisting of a copper build-up sheet, a gadolinium oxysulfide phosphor and an amorphous silicon detector for image formation. The MV-CBCT projection images were acquired with a CATPHAN 604 phantom and a Varian Truebeam linear accelerator at 2.5 MV, 6MV and 6MV FFF. The dose per projection was 0.01 MU, 0.017 and 0.05 for 2.5MV, 6MV and 6MV FFF, respectively. MVCBCT images were reconstructed with varying numbers of projections to provide a range of doses for evaluation. The HU accuracy and noise was assessed using the CATPHAN material inserts in the reconstructed images.
Results: HU values of insert materials were largely unaffected by reducing the projections number used for reconstruction. On average, less than 12HU difference was observed across different total MU values in any beam energy. Images reconstructed from 2.5MV beam acquisitions demonstrated the lowest noise (28.4HU at 7.2MU, 34.4HU at 3.6MU, 43.4HU at 1.8MU, 58HU at 0.9MU). The CNR of images obtained with 6MV FFF beam, was up to three times higher in materials with Zeff>5. However, considering the required MUs to achieve the CNR, the images from the 2.5MV beam demonstrated the best performance.
Conclusion: MVCBCT imaging with a novel MLI prototype mounted on a clinical linear accelerator demonstrated good HU and CNR performance along various reconstruction schemes used to achieve low delivered dose. Refinement of the reconstruction schemes with MLI specific corrections can further improve HU performance. The MLI could provide low-dose/high-accuracy images for adaptive radiotherapy applications.
Funding Support, Disclosures, and Conflict of Interest: The project described was supported, in part, by Award Number R01CA188446 from the National Cancer Institute. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.
Not Applicable / None Entered.
Not Applicable / None Entered.