Dynamic MRI Reconstruction Via Dual Deep Priors and Low-Rank Plus Sparse Modeling

Yongliang Sun, Siddhant Gautam, Chaoyan Huang, Nicole Seiberlich, Ismail Alkhouri, Saiprasad Ravishankar · 2026

Read Dynamic MRI Reconstruction Via Dual Deep Priors and Low-Rank Plus Sparse Modeling by Yongliang Sun, Siddhant Gautam, Chaoyan Huang, Nicole Seiberlich, Ismail Alkhouri, Saiprasad Ravishankar (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-18709.txt.

Description

Yongliang Sun and five co-authors propose, in a 2026 paper, a deep image prior method for dynamic MRI reconstruction. Learning methods need large sampled datasets and work poorly. Training-free prior methods adapt per scan but miss temporal structure and overfit. Their framework models space-time links through a low-rank plus sparse split, using two fresh networks jointly optimized by accelerated ADMM. The mix joins deep image prior rules with classical interpretability and convergence analysis. Across speedup factors, it beats classical, supervised, and unsupervised rebuilds.

Catalog record

CREATORYongliang Sun, Siddhant Gautam, Chaoyan Huang, Nicole Seiberlich, Ismail Alkhouri, Saiprasad Ravishankar
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE71937
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.18709

Browse the full library · About the Tower of Babel