发表机构
School of Mathematical Sciences, Capital Normal University; College of Mathematics Science, Inner Mongolia Normal University; National Center for Applied Mathematics Beijing, Capital Normal University; Academy for Multidisciplinary Studies, Capital Normal University（首都师范大学数学科学学院; 内蒙古师范大学数学科学学院; 首都师范大学北京应用数学中心; 首都师范大学交叉科学研究院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对X射线CT金属伪影问题，提出SCMA框架，通过结构条件化、金属感知损失及投影一致性校正提升伪影抑制与结构保留效果，实验验证其性能优于现有MAR方法。
AI 中文摘要
在X射线CT中，金属物体引发射线硬化、光子饥饿及散射，导致投影不一致、条纹、暗带和结构扭曲，损害临床诊断与定量分析。现有金属伪影减少（MAR）方法存在局限：基于优化的方法可能残留伪影或模糊结构，回归网络跨场景泛化性差，缺乏样本特异性结构引导与物理约束的生成模型易产生解剖结构不一致的结果。流匹配学习连续时间速度场，确定性地将源分布迁移至目标分布，为MAR提供灵活先验，但标准无约束流匹配未利用样本特异性结构、空间非均匀金属诱导退化或实测投影。为解决这些局限，我们提出SCMA——结构条件化与金属感知流匹配框架：首先，经线性插值校正的图像被输入速度网络，中间状态作为样本特异性结构条件，引导推理生成无伪影CT图像并保留解剖结构；其次，来自金属掩膜及其距离变换的时变空间权重被融入流匹配损失，以强化金属区域内部及周边的严重退化；最后，在推理过程中交替进行条件流匹配更新与投影一致性校正，使金属轨迹外的可靠测量值能约束预测结果。对模拟与真实CT数据的实验表明，与代表性MAR方法相比，SCMA能更有效抑制金属伪影、保留局部解剖结构，并减少与投影测量不一致的幻觉式结构。
英文摘要
In X-ray CT, metallic objects cause beam hardening, photon starvation, and scattering, leading to projection inconsistency, streaks, dark bands, and structural distortions that compromise clinical diagnosis and quantitative analysis. Existing metal artifact reduction (MAR) methods remain limited: optimization-based methods may leave residual artifacts or blur structures, regression networks may generalize poorly across scenarios, and generative models without sample-specific structural guidance and physical constraints may produce anatomically inconsistent structures. Flow Matching learns a continuous-time velocity field that deterministically transports a source distribution to a target distribution, providing a flexible MAR prior. However, standard unconditional Flow Matching does not exploit sample-specific structure, spatially nonuniform metal-induced degradation, or measured projections. To address these limitations, we propose SCMA, a structure-conditioned and metal-aware Flow Matching framework. First, a linear-interpolation-corrected image is fed into the velocity network with the intermediate state as a sample-specific structural condition, guiding inference toward artifact-free CT images while preserving anatomy. Second, time-varying spatial weights from the metal mask and its distance transform are incorporated into the Flow Matching loss to emphasize severe degradation within and around metal regions. Finally, conditional Flow Matching updates alternate with projection-consistency correction during inference, allowing reliable measurements outside metal traces to constrain predictions. Experiments on simulated and real CT data demonstrate that SCMA more effectively suppresses metal artifacts, preserves local anatomical structures, and reduces hallucination-like structures inconsistent with projection measurements than representative MAR methods.
