Efficient Vision Mamba for MRI Super-Resolution via Hybrid Selective Scanning

2025-12-22

AI Summary

High-resolution MRI is clinically limited by prolonged acquisition times, while existing deep learning super-resolution methods struggle to balance reconstruction fidelity and computational efficiency. To address this, we propose a lightweight, efficient medical MRI super-resolution framework. Our method introduces a novel hybrid selective scanning mechanism and a multi-head selective state space model (MHSSM), forming the first anatomy-aware MambaFormer module—integrating 2D patch modeling, depthwise separable convolutions, and a gated channel-mixing mechanism, augmented by a lightweight channel-wise MLP. Evaluated on 7T brain and 1.5T prostate MRI datasets, our approach achieves SSIM scores of 0.951 and 0.770, respectively, with only 0.9M parameters and 97.5% lower FLOPs than mainstream methods (e.g., SwinIR, MambaIR, and diffusion-based models). This substantial improvement in efficiency–fidelity trade-off significantly enhances clinical deployability.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Intelligent Robots: Multi-Robot Systems

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Background: High-resolution MRI is critical for diagnosis, but long acquisition times limit clinical use. Super-resolution (SR) can enhance resolution post-scan, yet existing deep learning methods face fidelity-efficiency trade-offs. Purpose: To develop a computationally efficient and accurate deep learning framework for MRI SR that preserves anatomical detail for clinical integration. Materials and Methods: We propose a novel SR framework combining multi-head selective state-space models (MHSSM) with a lightweight channel MLP. The model uses 2D patch extraction with hybrid scanning to capture long-range dependencies. Each MambaFormer block integrates MHSSM, depthwise convolutions, and gated channel mixing. Evaluation used 7T brain T1 MP2RAGE maps (n=142) and 1.5T prostate T2w MRI (n=334). Comparisons included Bicubic interpolation, GANs (CycleGAN, Pix2pix, SPSR), transformers (SwinIR), Mamba (MambaIR), and diffusion models (I2SB, Res-SRDiff). Results: Our model achieved superior performance with exceptional efficiency. For 7T brain data: SSIM=0.951+-0.021, PSNR=26.90+-1.41 dB, LPIPS=0.076+-0.022, GMSD=0.083+-0.017, significantly outperforming all baselines (p<0.001). For prostate data: SSIM=0.770+-0.049, PSNR=27.15+-2.19 dB, LPIPS=0.190+-0.095, GMSD=0.087+-0.013. The framework used only 0.9M parameters and 57 GFLOPs, reducing parameters by 99.8% and computation by 97.5% versus Res-SRDiff, while outperforming SwinIR and MambaIR in accuracy and efficiency. Conclusion: The proposed framework provides an efficient, accurate MRI SR solution, delivering enhanced anatomical detail across datasets. Its low computational demand and state-of-the-art performance show strong potential for clinical translation.

Problem

• Develops efficient deep learning for MRI super-resolution to preserve anatomical details.
• Addresses fidelity-efficiency trade-offs in existing methods for clinical integration.
• Proposes a lightweight framework combining selective state-space models and channel MLP.

Innovation

• Hybrid selective scanning captures long-range dependencies efficiently
• Multi-head selective state-space models integrate with lightweight channel MLP
• MambaFormer blocks combine MHSSM, depthwise convolutions, and gated mixing