用3D高斯点云重构多层磁共振图像，速度快质量高。
M-Gaussian: An Magnetic Gaussian Framework for Efficient Multi-Stack MRI Reconstruction
- 用物理一致的磁性高斯体素实现图像渲染
- 在FeTA数据集上达40.31 dB PSNR，速度提升14倍
- 适合临床快速高分辨成像，尤其胎儿脑部扫描
磁共振成像（MRI）是重要的无创影像技术。临床中常采用多层厚层采集以缩短扫描时间并降低运动伪影，尤其适用于胎儿脑成像等挑战场景。但由此产生的严重层面间各向异性会损害体积分析与定量评估，亟需重建各向同性的高分辨率体积。隐式神经表示虽能实现高质量重建，但因网络结构复杂导致计算效率低下。本文提出M-Gaussian，将3D高斯点云渲染适配至MRI重建。贡献包括：(1) 物理一致的磁性高斯基元与体渲染方法；(2) 神经残差场用于高频细节优化；(3) 多分辨率渐进训练策略。该方法在质量与速度间取得最佳平衡。在FeTA数据集上，达到40.31 dB PSNR，同时比基准快14倍，首次成功将3D高斯点云应用于多层MRI重建。
原文摘要 · Abstract (English)
Magnetic Resonance Imaging (MRI) is a crucial non-invasive imaging modality. In routine clinical practice, multi-stack thick-slice acquisitions are widely used to reduce scan time and motion sensitivity, particularly in challenging scenarios such as fetal brain imaging. However, the resulting severe through-plane anisotropy compromises volumetric analysis and downstream quantitative assessment, necessitating robust reconstruction of isotropic high-resolution volumes. Implicit neural representation methods, while achieving high quality, suffer from computational inefficiency due to complex network structures. We present M-Gaussian, adapting 3D Gaussian Splatting to MRI reconstruction. Our contributions include: (1) Magnetic Gaussian primitives with physics-consistent volumetric rendering, (2) neural residual field for high-frequency detail refinement, and (3) multi-resolution progressive training. Our method achieves an optimal balance between quality and speed. On the FeTA dataset, M-Gaussian achieves 40.31 dB PSNR while being 14 times faster, representing the first successful adaptation of 3D Gaussian Splatting to multi-stack MRI reconstruction.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。