PAS-Mamba: Phase-Amplitude-Spatial State Space Model for MRI Reconstruction

2026-01-20

AI Summary

This work addresses the limitation of existing MRI reconstruction methods that fail to differentiate phase and magnitude information in the frequency domain, leading to feature coupling and degraded reconstruction quality. To overcome this, we propose PAS-Mamba, a novel framework that explicitly decouples phase and magnitude modeling in the frequency domain for the first time in MRI reconstruction. Our approach introduces a circular frequency-domain scanning strategy to serialize spectral features while incorporating LocalMamba to preserve spatial locality. Furthermore, we design a dual-domain complementary fusion module that adaptively integrates the decoupled frequency-domain features with spatial-domain information. Extensive experiments demonstrate that PAS-Mamba significantly outperforms state-of-the-art methods on both the IXI and fastMRI knee datasets, achieving superior reconstruction accuracy and structural fidelity.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Feature Construction/Reformulation
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Joint feature modeling in both the spatial and frequency domains has become a mainstream approach in MRI reconstruction. However, existing methods generally treat the frequency domain as a whole, neglecting the differences in the information carried by its internal components. According to Fourier transform theory, phase and amplitude represent different types of information in the image. Our spectrum swapping experiments show that magnitude mainly reflects pixel-level intensity, while phase predominantly governs image structure. To prevent interference between phase and magnitude feature learning caused by unified frequency-domain modeling, we propose the Phase-Amplitude-Spatial State Space Model (PAS-Mamba) for MRI Reconstruction, a framework that decouples phase and magnitude modeling in the frequency domain and combines it with image-domain features for better reconstruction. In the image domain, LocalMamba preserves spatial locality to sharpen fine anatomical details. In frequency domain, we disentangle amplitude and phase into two specialized branches to avoid representational coupling. To respect the concentric geometry of frequency information, we propose Circular Frequency Domain Scanning (CFDS) to serialize features from low to high frequencies. Finally, a Dual-Domain Complementary Fusion Module (DDCFM) adaptively fuses amplitude phase representations and enables bidirectional exchange between frequency and image domains, delivering superior reconstruction. Extensive experiments on the IXI and fastMRI knee datasets show that PAS-Mamba consistently outperforms state of the art reconstruction methods.

Problem

• MRI reconstruction
• phase-amplitude decoupling
• frequency domain modeling
• spatial-frequency joint modeling
• image reconstruction

Innovation

• Phase-Amplitude Decoupling
• State Space Model
• MRI Reconstruction
• Circular Frequency Domain Scanning
• Dual-Domain Fusion