Harnessing Magnitude-Only and Complex Measurements for Improved Dynamic MRI Reconstruction with Learned Priors
Abstract
MRI reconstruction methods for undersampled k-space data naturally utilize complex-valued measurements. Parallel developments in sparse phase retrieval have shown that magnitude-only measurements may provide complementary information for signal recovery. However, their use in MRI reconstruction remains largely unexplored, due to lack of practical settings where informative magnitude measurements can be obtained without additional scan time. In this work, we investigate the use of auxiliary k-space magnitude information for accelerated steady-state dynamic MRI reconstruction, and demonstrate strong consistency of k-space magnitudes across time-frames. Building on this observation, we propose ℂ+Mag, a magnitude-informed physics-driven deep learning reconstruction method. The proposed method employs an ADMM-based unrolling framework with a novel magnitude-aware data-fidelity formulation, where quadratically smoothed optimization and momentum-based updates are introduced to address the non-differentiability and non-convexity of the magnitude constraints. Experiments on retrospectively undersampled cine MRI and phase-contrast flow MRI datasets, as well as prospectively undersampled real-time cine MRI acquisitions, demonstrate improved artifact suppression, sharper anatomical recovery, and better preservation of phase information compared to conventional PD-DL methods, which is further supported through blinded expert reader evaluations.
Versions
Related
- Harnessing Post-selected von Neumann Measurements to Strengthen Entanglement and Quantum Signatures in Coherent States
- Evidence for the decay B0 → μ+ μ- and improved measurements of the decay B0s → μ+ μ-
- Improved Radiofrequency Safety Modelling in MRI Using In Vivo Measurements of Brain Conductivity
- From Tunnel Overburden Reconstruction to Tidal Measurements via atmospheric muon
- Observation of χcJ→ ΛK0SΞ̅0 + c.c. and improved measurements of χcJ→ ΛK-Ξ̅+ + c.c.
BibTeX
@misc{saberi2026harnessing,
title = {{Harnessing Magnitude-Only and Complex Measurements for Improved Dynamic MRI Reconstruction with Learned Priors}},
author = {Mahdi Saberi and Yaşar Utku Alçalar and Merve Gülle and Chetan Shenoy and Mehmet Akçakaya},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2608.18036},
eprint = {2608.18036},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2608.18036},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Harnessing Magnitude-Only and Complex Measurements for Improved Dynamic MRI Reconstruction with Learned Priors
AU - Mahdi Saberi
AU - Yaşar Utku Alçalar
AU - Merve Gülle
AU - Chetan Shenoy
AU - Mehmet Akçakaya
PY - 2026
DA - 2026/10/20
PB - arXiv
DO - 10.48550/arxiv.2608.18036
UR - https://arxiv.org/abs/2608.18036
AB - MRI reconstruction methods for undersampled k-space data naturally utilize complex-valued measurements. Parallel developments in sparse phase retrieval have shown that magnitude-only measurements may provide complementary information for signal recovery. However, their use in MRI reconstruction remains largely unexplored, due to lack of practical settings where informative magnitude measurements can be obtained without additional scan time. In this work, we investigate the use of auxiliary k-space magnitude information for accelerated steady-state dynamic MRI reconstruction, and demonstrate strong consistency of k-space magnitudes across time-frames. Building on this observation, we propose ℂ+Mag, a magnitude-informed physics-driven deep learning reconstruction method. The proposed method employs an ADMM-based unrolling framework with a novel magnitude-aware data-fidelity formulation, where quadratically smoothed optimization and momentum-based updates are introduced to address the non-differentiability and non-convexity of the magnitude constraints. Experiments on retrospectively undersampled cine MRI and phase-contrast flow MRI datasets, as well as prospectively undersampled real-time cine MRI acquisitions, demonstrate improved artifact suppression, sharper anatomical recovery, and better preservation of phase information compared to conventional PD-DL methods, which is further supported through blinded expert reader evaluations.
ER -