基于多技术融合的CT金属伪影去除方法研究

Research on CT Metal Artifact Removal Method Based on Multi Technology Fusion

摘要: 在计算机断层扫描(computed tomography, CT)检查中，患者体内留存的金属植入物(如人工关节假体、颅内动脉瘤栓塞弹簧圈、口腔修复用金属冠桥等)易诱发多种类型的金属伪影，显著降低图像的密度分辨率与空间分辨率，干扰病灶的精准识别及术后疗效的客观评估。金属伪影的成因复杂，是光子饥饿效应、X线束硬化效应、散射线干扰等多因素共同作用的结果。如何有效抑制金属伪影，已成为提升CT诊断效能的关键课题。当前单一技术难以兼顾伪影抑制效果、图像分辨率及辐射剂量需求，多技术融合已成为该领域的核心研究方向。本文系统梳理当前临床主流及前沿的CT金属伪影减轻技术，深入阐述双能量成像、迭代重建算法等核心技术的作用机制，对比各类技术的优势与局限性，重点探讨多技术融合的应用模式与优化路径，并结合最新研究进展对该领域的发展趋势进行展望，旨在为临床精准成像提供理论参考与实践指导。

Abstract: In computed tomography (CT) scans, metal implants left in the patient’s body (such as artificial joint prostheses, intracranial aneurysm embolization coils, metal crown bridges for oral restoration, etc.) are prone to induce various types of metal artifacts, significantly reducing the density resolution and spatial resolution of the image, interfering with the accurate identification of lesions and the objective evaluation of postoperative efficacy. The causes of metal artifacts are complex and are the result of multiple factors such as photon starvation effect, X-ray beam hardening effect, and scattered line interference. How to effectively suppress metal artifacts has become a key issue in improving the diagnostic efficiency of CT. The current single technology is difficult to balance the requirements of artifact suppression, image resolution, and radiation dose. Multi technology fusion has become the core research direction in this field. This article systematically reviews the current mainstream and cutting-edge CT metal artifact reduction technologies in clinical practice, elaborates on the mechanisms of core technologies such as dual energy imaging and iterative reconstruction algorithms, compares the advantages and limitations of various technologies, focuses on exploring the application mode and optimization path of multi technology fusion, and prospects the development trend of this field in combination with the latest research progress, aiming to provide theoretical reference and practical guidance for clinical precision imaging.