Spatiotemporal Diffusion Model with Paired Sampling for Accelerated Cardiac Cine MRI

TL;DR

This paper introduces a spatiotemporal diffusion model with paired sampling to enhance the quality of accelerated cardiac cine MRI, achieving sharper images and clearer motion delineation under high undersampling conditions.

Contribution

The study presents a novel diffusion model combined with a paired sampling strategy that significantly improves image sharpness and reduces noise in accelerated cardiac MRI reconstructions.

Findings

Sharper tissue boundaries achieved

Clearer motion delineation confirmed by experts

Reduced artificial noise in generated images

Abstract

Current deep learning reconstruction for accelerated cardiac cine MRI suffers from spatial and temporal blurring. We aim to improve image sharpness and motion delineation for cine MRI under high undersampling rates. A spatiotemporal diffusion enhancement model conditional on an existing deep learning reconstruction along with a novel paired sampling strategy was developed. The diffusion model provided sharper tissue boundaries and clearer motion than the original reconstruction in experts evaluation on clinical data. The innovative paired sampling strategy substantially reduced artificial noises in the generative results.