Adaptive Gate-Aware Mamba Networks for Magnetic Resonance Fingerprinting

arXiv:2507.03369 · eess.IV, cs.LG, eess.SP · Submitted 2025-07-04 · Read on arXiv

eess.IV, cs.LG, eess.SP

Submitted: 2025-07-04

Updated: 2025-07-04

Comments: 31 pages, 7 figures

Journal ref: 2026 IEEE 23rd International Symposium on Biomedical Imaging (ISBI), 2026

DOI: 10.1109/ISBI61048.2026.11515932

License: http://creativecommons.org/licenses/by/4.0/

The gist: Magnetic Resonance Fingerprinting (MRF) enables fast quantitative imaging by matching signal evolutions to a predefined dictionary.

Terminology

Abstract

Magnetic Resonance Fingerprinting (MRF) enables fast quantitative imaging by matching signal evolutions to a predefined dictionary. However, conventional dictionary matching suffers from exponential growth in computational cost and memory usage as the number of parameters increases, limiting its scalability to multi-parametric mapping. To address this, recent work has explored deep learning-based approaches as alternatives to DM. We propose GAST-Mamba, an end-to-end framework that combines a dual Mamba-based encoder with a Gate-Aware Spatial-Temporal (GAST) processor. Built on structured state-space models, our architecture efficiently captures long-range spatial dependencies with linear complexity. On 5 times accelerated simulated MRF data (200 frames), GAST-Mamba achieved a T1 PSNR of 33.12 dB, outperforming SCQ (31.69 dB). For T2 mapping, it reached a PSNR of 30.62 dB and SSIM of 0.9124. In vivo experiments further demonstrated improved anatomical detail and reduced artifacts. Ablation studies confirmed that each component contributes to performance, with the GAST module being particularly important under strong undersampling. These results demonstrate the effectiveness of GAST-Mamba for accurate and robust reconstruction from highly undersampled MRF acquisitions, offering a scalable alternative to traditional DM-based methods.

Sources

Related papers