Measurement-Conditioned Denoising Diffusion Probabilistic Model for Under-Sampled Medical Image Reconstruction

We propose a novel and unified method, measurement-conditioned denoising diffusion probabilistic model (MC-DDPM), for under-sampled medical image reconstruction based on DDPM. Different from previous works, MC-DDPM is defined in measurement domain (e.g. k-space MRI reconstruction) conditioned under-sampling mask. apply this method to accelerate the experimental results show excellent performance, outperforming full supervision baseline state-of-the-art score-based method. Due its generative nature, can also quantify uncertainty of reconstruction. Our code available github ( https://github.com/Theodore-PKU/MC-DDPM ).