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

TL;DR

This paper introduces a data-free dynamic MRI reconstruction method that combines deep image priors with low-rank plus sparse modeling, effectively capturing spatiotemporal correlations without large training datasets.

Contribution

It proposes a structured DIP framework with a novel eADMM optimization that models spatiotemporal dynamics via low-rank and sparse components, with convergence guarantees.

Findings

Outperforms classical and existing deep learning MRI methods across various acceleration factors.

Effectively models spatiotemporal correlations without large training datasets.

Provides theoretical convergence analysis for the proposed nonconvex optimization.

Abstract

Dynamic MRI reconstruction from undersampled measurements is a challenging inverse problem that requires preserving both spatial reconstruction quality and temporal consistency across the frames of the cine series. While recent learning-based approaches achieve strong performance, they heavily rely on large training, mostly fully sampled, datasets, and may otherwise generalize poorly. In contrast, training-data-free methods such as deep image prior (DIP) adapt directly to individual scans but often fail to fully exploit temporal structure and are prone to overfitting. They are particularly attractive for dynamic MRI due to the limited large, public, high-quality datasets. In this work, we propose a structured DIP framework for dynamic MRI reconstruction that explicitly models spatiotemporal correlations through a low-rank plus sparse (L+S) decomposition. Instead of directly reconstructing the cine image series, we parameterize the low-rank background and sparse dynamic components using two DIP untrained convolutional neural networks, jointly optimized using accelerated extrapolated ADMM (eADMM). This formulation combines the implicit regularization of DIP with the interpretability of classical L+S regularization. We provide a convergence analysis for the proposed eADMM algorithm in the presence of DIP-based nonconvex parameterizations. In particular, we establish a sufficient descent property and show that every cluster point of the generated sequence is a critical point of the associated Lyapunov function. Across various acceleration factors, our numerical results demonstrate that the proposed method consistently outperforms classical reconstruction and existing supervised and unsupervised MRI reconstruction techniques.