用确定性流匹配加速低视角CT重建，效率提升显著。
Efficient Flow Matching for Sparse-View CT Reconstruction
- 采用确定性流匹配替代随机扩散过程，避免噪声干扰数据一致性校正。
- 通过复用相邻步的预测速度场，将神经网络评估次数减少超50%。
- 适合临床急用场景，兼顾重建质量与推理速度，代码开源。
生成模型，尤其是扩散模型（DM），在解决病态逆问题的计算机断层扫描（CT）重建中展现出强大潜力，可作为表达性强的先验。然而，基于扩散的重建依赖随机微分方程（SDE）进行前向扩散和反向去噪，其中随机性可能干扰CT重建中反复进行的数据一致性修正。由于临床与介入场景对重建速度要求高，提升效率至关重要。相比之下，流匹配（FM）模型以确定性常微分方程（ODE）采样，轨迹平滑且无随机噪声注入，天然适配重复数据一致性操作。我们观察到，FM预测的速度场在相邻步骤间具有强相关性。受此启发，提出基于FM的CT重建框架（FMCT）及其高效变体（EFMCT），通过在连续步骤中复用先前预测的速度场，显著减少神经网络函数评估次数（NFEs），从而提升推理效率。理论分析表明，在结合数据一致性操作时，速度复用引入的误差是可控的。大量实验显示，FMCT/EFMCT在保持竞争性重建质量的同时，相较扩散方法大幅提升了计算效率。代码已开源：https://github.com/EFMCT/EFMCT。
原文摘要 · Abstract (English)
Generative models, particularly Diffusion Models (DM), have shown strong potential for Computed Tomography (CT) reconstruction serving as expressive priors for solving ill-posed inverse problems. However, diffusion-based reconstruction relies on Stochastic Differential Equations (SDEs) for forward diffusion and reverse denoising, where such stochasticity can interfere with repeated data consistency corrections in CT reconstruction. Since CT reconstruction is often time-critical in clinical and interventional scenarios, improving reconstruction efficiency is essential. In contrast, Flow Matching (FM) models sampling as a deterministic Ordinary Differential Equation (ODE), yielding smooth trajectories without stochastic noise injection. This deterministic formulation is naturally compatible with repeated data consistency operations. Furthermore, we observe that FM-predicted velocity fields exhibit strong correlations across adjacent steps. Motivated by this, we propose an FM-based CT reconstruction framework (FMCT) and an efficient variant (EFMCT) that reuses previously predicted velocity fields over consecutive steps to substantially reduce the number of Neural network Function Evaluations (NFEs), thereby improving inference efficiency. We provide theoretical analysis showing that the error introduced by velocity reuse is bounded when combined with data consistency operations. Extensive experiments demonstrate that FMCT/EFMCT achieve competitive reconstruction quality while significantly improving computational efficiency compared with diffusion-based methods. The codebase is open-sourced at https://github.com/EFMCT/EFMCT.
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。