Efficient vision mamba for MRI super-resolution via hybrid selective scanning.

BACKGROUND: High-resolution MRI is essential for accurate diagnosis and treatment planning, but its clinical acquisition is often constrained by long scanning times, which increase patient discomfort and reduce scanner throughput. While super-resolution (SR) techniques offer a post-acquisition solution to enhance resolution, existing deep learning approaches face trade-offs between reconstruction fidelity and computational efficiency, limiting their clinical applicability.

PURPOSE: This study aims to develop an efficient and accurate deep learning framework for MRI SR that preserves fine anatomical detail while maintaining low computational overhead, enabling practical integration into clinical workflows.

MATERIALS AND METHODS: We propose a novel SR framework based on multi-head selective state-space models (MHSSM) integrated with a lightweight channel multilayer perceptron (MLP). The model employs 2D patch extraction with hybrid scanning strategies (vertical, horizontal, and diagonal) to capture long-range dependencies while mitigating pixel forgetting. Each MambaFormer block combines MHSSM, depthwise convolutions, and gated channel mixing to balance local and global feature representation. The framework was trained and evaluated on two distinct datasets: 7T brain T1 MP2RAGE maps (142 subjects) and 1.5T prostate T2w MRI (334 subjects). Performance was compared against multiple baselines including Bicubic interpolation, GAN-based (CycleGAN, Pix2pix, SPSR), transformer-based (SwinIR), Mamba-based (MambaIR), and diffusion-based (I2SB, Res-SRDiff) methods.

RESULTS: The proposed model demonstrated superior performance across all evaluation metrics while maintaining exceptional computational efficiency. On the 7T brain dataset, our method achieved the highest structural similarity (SSIM: 0.951 ± 0.021 $0.951 \pm 0.021$ ) and peak signal-to-noise ratio (PSNR: 26.90 ± 1.41 $26.90 \pm 1.41$ dB), along with the best perceptual quality scores (LPIPS: 0.076 ± 0.022 $0.076 \pm 0.022$ ; GMSD: 0.083 ± 0.017 $0.083 \pm 0.017$ ). These results represented statistically significant improvements over all baselines ( p < 0.001 $p < 0.001$ ), including a 2.1% SSIM gain over SPSR and a 2.4% PSNR improvement over Res-SRDiff. For the prostate dataset, the model similarly outperformed competing approaches, achieving SSIM of 0.770 ± 0.049 $0.770 \pm 0.049$ , PSNR of 27.15 ± 2.19 $27.15 \pm 2.19$ dB, LPIPS of 0.190 ± 0.095 $0.190 \pm 0.095$ , and GMSD of 0.087 ± 0.013 $0.087 \pm 0.013$ . Notably, our framework accomplished these results with only 0.9 million parameters and 57 GFLOPs, representing reductions of 99.8% in parameters and 97.5% in computational operations compared to Res-SRDiff, while also substantially outperforming SwinIR and MambaIR in both accuracy and efficiency metrics.

CONCLUSION: The proposed framework provides a computationally efficient yet accurate solution for MRI SR, delivering well-defined anatomical details and improved perceptual fidelity across anatomically distinct datasets. By significantly reducing computational demands while maintaining state-of-the-art performance, the model offers strong potential for feasibility toward clinical translation and scalable integration into future imaging workflows.