Saved in:
Bibliographic Details
Main Authors: Fang, Pengcheng, Chen, Hongli, Yao, Guangzhen, Shi, Jian, Tang, Fangfang, Cai, Xiaohao, Shan, Shanshan, Liu, Feng
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.18534
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917100145082368
author Fang, Pengcheng
Chen, Hongli
Yao, Guangzhen
Shi, Jian
Tang, Fangfang
Cai, Xiaohao
Shan, Shanshan
Liu, Feng
author_facet Fang, Pengcheng
Chen, Hongli
Yao, Guangzhen
Shi, Jian
Tang, Fangfang
Cai, Xiaohao
Shan, Shanshan
Liu, Feng
contents Reconstructing high-fidelity MR images from undersampled k-space data requires recovering high-frequency details while maintaining anatomical coherence. We present HiFi-MambaV2, a hierarchical shared-routed Mixture-of-Experts (MoE) Mamba architecture that couples frequency decomposition with content-adaptive computation. The model comprises two core components: (i) a separable frequency-consistent Laplacian pyramid (SF-Lap) that delivers alias-resistant, stable low- and high-frequency streams; and (ii) a hierarchical shared-routed MoE that performs per-pixel top-1 sparse dispatch to shared experts and local routers, enabling effective specialization with stable cross-depth behavior. A lightweight global context path is fused into an unrolled, data-consistency-regularized backbone to reinforce long-range reasoning and preserve anatomical coherence. Evaluated on fastMRI, CC359, ACDC, M4Raw, and Prostate158, HiFi-MambaV2 consistently outperforms CNN-, Transformer-, and prior Mamba-based baselines in PSNR, SSIM, and NMSE across single- and multi-coil settings and multiple acceleration factors, consistently surpassing consistent improvements in high-frequency detail and overall structural fidelity. These results demonstrate that HiFi-MambaV2 enables reliable and robust MRI reconstruction.
format Preprint
id arxiv_https___arxiv_org_abs_2511_18534
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle HiFi-MambaV2: Hierarchical Shared-Routed MoE for High-Fidelity MRI Reconstruction
Fang, Pengcheng
Chen, Hongli
Yao, Guangzhen
Shi, Jian
Tang, Fangfang
Cai, Xiaohao
Shan, Shanshan
Liu, Feng
Computer Vision and Pattern Recognition
Reconstructing high-fidelity MR images from undersampled k-space data requires recovering high-frequency details while maintaining anatomical coherence. We present HiFi-MambaV2, a hierarchical shared-routed Mixture-of-Experts (MoE) Mamba architecture that couples frequency decomposition with content-adaptive computation. The model comprises two core components: (i) a separable frequency-consistent Laplacian pyramid (SF-Lap) that delivers alias-resistant, stable low- and high-frequency streams; and (ii) a hierarchical shared-routed MoE that performs per-pixel top-1 sparse dispatch to shared experts and local routers, enabling effective specialization with stable cross-depth behavior. A lightweight global context path is fused into an unrolled, data-consistency-regularized backbone to reinforce long-range reasoning and preserve anatomical coherence. Evaluated on fastMRI, CC359, ACDC, M4Raw, and Prostate158, HiFi-MambaV2 consistently outperforms CNN-, Transformer-, and prior Mamba-based baselines in PSNR, SSIM, and NMSE across single- and multi-coil settings and multiple acceleration factors, consistently surpassing consistent improvements in high-frequency detail and overall structural fidelity. These results demonstrate that HiFi-MambaV2 enables reliable and robust MRI reconstruction.
title HiFi-MambaV2: Hierarchical Shared-Routed MoE for High-Fidelity MRI Reconstruction
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.18534