SO-Mamba: State-Ownership Mamba for Unrolled MRI Reconstruction

2026-05-21

AI Summary

Accelerated MRI reconstruction under undersampled conditions demands a delicate balance between fine detail recovery and anatomical consistency, yet existing methods struggle to distinguish stable content to retain from information requiring updates during iterative refinement. This work proposes SO-Mamba, an unfolded reconstruction framework based on state space models, which introduces a novel State Ownership Routing (SOR) mechanism to explicitly separate recursively persistent content from stage-specific evidence. Precise control is achieved through affine-modulated state interfaces, output correction pathways, and selective scan trajectory analysis. Evaluated across five diverse MRI benchmarks encompassing varied anatomies, sampling patterns, and coil configurations, SO-Mamba consistently outperforms CNN-, Transformer-, and Mamba-based baselines while maintaining computational efficiency.

Abstract

Accelerated MRI reconstruction requires recovering missing details while preserving anatomically coherent structures across large spatial regions. State-space models such as Mamba provide efficient long-range modeling, making them attractive learned regularizers for unrolled reconstruction. However, in a data-consistency-coupled unrolled solver, different stages operate on different reconstruction iterates, where the resident carrier should preserve coherent reconstruction content across stages while stage-dependent non-resident evidence is tied to the current update. Treating these roles uniformly can place persistent resident-carrier evidence and update-dependent non-resident evidence into the same recurrent content route. We therefore propose SO-Mamba, a state-ownership Mamba regularizer that assigns reconstruction evidence within each Mamba stage to recurrent residency, state-interface access, and non-state output correction. SO-Mamba implements this ownership rule with a State-Ownership Router (SOR), which constructs a resident carrier for recurrent content and routes non-resident evidence to affine modulation of the B/C state interfaces and an output correction outlet. The resident carrier supplies the Mamba content route, while the non-resident evidence stream adapts the state interfaces and contributes through the output outlet without entering the recurrent content route. We further introduce a two-level outer-band leakage diagnostic that separates hidden-state storage from readout expression by measuring outer-band energy in the selective-scan state trajectory and the post-scan Mamba readout. Experiments on five public MRI reconstruction benchmarks spanning diverse anatomies, sampling patterns, and coil configurations show that SO-Mamba consistently improves over CNN-, Transformer-, and Mamba-based baselines with competitive computational efficiency.

Problem

• Accelerated MRI reconstruction
• state-space models
• unrolled reconstruction
• recurrent content routing
• anatomical coherence

Innovation

• State-Ownership Mamba
• unrolled MRI reconstruction
• state-space models
• recurrent content routing
• affine modulation