Total: 1
Photon-counting computed tomography (PCCT) has emerged as a promising imaging technique, enabling spectral imaging and material decomposition (MD). However, images typically suffer from a low signal-to-noise ratio due to constraints such as low photon counts and sparse-view settings. Variational methods minimize a data-fit function coupled with handcrafted regularizers but are highly dependent on the choice of the regularizers. Artificial intelligence (AI)-based approaches and more particularly convolutional neural networks (CNNs) are now considered the state-of-the-art approach and can be used as an end-to-end method for MD or to implicitly learn an a priori. In the last few years, diffusion models (DMs) became predominant in the field of generative models where a distribution function is learned. This distribution function can be used as a prior for solving inverse problems. This work investigates the use of DMs as regularizers for MD tasks in PCCT. MD by diffusion posterior sampling (DPS) can be achieved. Three DPS-based approaches -- image-domain two-step DPS (im-TDPS), projection-domain two-step DPS (proj-TDPS), and one-step DPS (ODPS) -- are evaluated. The first two methods perform MD in two steps: im-TDPS samples spectral images by DPS then performs image-based MD, while proj-TDPS performs projection-based MD then samples material images by DPS. The last method, ODPS, samples the material images directly from the measurement data. The results indicate that ODPS achieves superior performance compared to im-TDPS and proj-TDPS, providing sharper, noise-free and crosstalk-free images. Furthermore, we introduce a novel hybrid ODPS method combining DM priors with standard variational regularizers for scenarios involving materials absent from the training dataset. This hybrid method demonstrates improved material reconstruction quality compared to a standard variational method.