Adaptive Gate-Aware Mamba Networks for Magnetic Resonance Fingerprinting

TL;DR

This paper introduces GAST-Mamba, a novel deep learning framework for Magnetic Resonance Fingerprinting that efficiently captures spatial dependencies, enabling accurate, scalable, and artifact-reduced imaging from highly undersampled data.

Contribution

GAST-Mamba combines structured state-space models with a gate-aware processor, providing a scalable, end-to-end solution for MRF reconstruction that outperforms traditional dictionary matching.

Findings

Achieved higher PSNR and SSIM compared to baseline methods.

Demonstrated robustness on accelerated simulated and in vivo MRF data.

Component ablation confirmed the importance of the GAST module.

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.