Spatiotemporal Diffusion Model with Paired Sampling for Accelerated Cardiac Cine MRI

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Qiu, Shihan, Pan, Shaoyan, Liu, Yikang, Zhao, Lin, Xu, Jian, Liu, Qi, Chen, Terrence, Chen, Eric Z., Chen, Xiao, Sun, Shanhui
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866911796309262336
author Qiu, Shihan
Pan, Shaoyan
Liu, Yikang
Zhao, Lin
Xu, Jian
Liu, Qi
Chen, Terrence
Chen, Eric Z.
Chen, Xiao
Sun, Shanhui
author_facet Qiu, Shihan
Pan, Shaoyan
Liu, Yikang
Zhao, Lin
Xu, Jian
Liu, Qi
Chen, Terrence
Chen, Eric Z.
Chen, Xiao
Sun, Shanhui
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2403_08758
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Spatiotemporal Diffusion Model with Paired Sampling for Accelerated Cardiac Cine MRI
Qiu, Shihan
Pan, Shaoyan
Liu, Yikang
Zhao, Lin
Xu, Jian
Liu, Qi
Chen, Terrence
Chen, Eric Z.
Chen, Xiao
Sun, Shanhui
Image and Video Processing
Computer Vision and Pattern Recognition
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.
title Spatiotemporal Diffusion Model with Paired Sampling for Accelerated Cardiac Cine MRI
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2403.08758