Diffusion model-based single-scan dual-energy cone-beam CT using primary beam splitting

Dual-energy cone-beam CT (DECBCT) has great potential for quantitative imaging tasks in CBCT-guided radiation therapy. However, the lack of a practical single-scan solution in data acquisition impedes the practice of CBCT-based adaptive radiotherapy (ART). In this work, we propose an efficient way to achieve DECBCT using the primary beam splitting in a single short scan or half-fan scan. To restore complete dual-energy sinograms sufficient for analytical image reconstruction, the conditional diffusion model is introduced to convert the acquired spectral-mixed sinogram to the complete dual-energy sinograms via a data refinement strategy. The proposed method is compared with the other two diffusion model-based methods and the preliminary results demonstrate the feasibility of the proposed method in quantitative imaging tasks, making it a promising solution to CBCT-based dose calculation and replanning.