ISSN 1004-4140
    CN 11-3017/P

    基于光子计数CT多能量成像联合iMAR技术在抑制心脏起搏器金属伪影中的应用

    Combining Iterative Metal Artifact Reduction and Virtual Monoenergetic Images to Reduce Pacemaker-associated Artifacts in Photon-counting Computed Tomography CT

    • 摘要: 目的:评估光子计数CT(PCCT)虚拟单能量图像(VMI)结合迭代金属伪影去除(iMAR)技术,在抑制心脏起搏器金属伪影中的可行性,并确定最优重建能级。材料与方法:前瞻性纳入2024年2月至12月期间接受胸部PCCT扫描的42例心脏起搏器植入患者。图像分别重建VMI:70~150 keV(间隔20 keV),并采用联合或不联合iMAR的条件下。采用Friedman检验分别比较伪影指数(AIX)和校正衰减(△HU)之间的差异,主观评分一致性评估采用加权Kappa分析。主观评分比较采用单因素重复测量ANOVA。结果:与未联合iMAR组相比,采用iMAR技术组,所有伪影类型包括高/低密度肌肉伪影区、(锁)骨、血管及甲状腺伪影区的AIX值均显著降低。在110 keV+iMAR重建组观察到最优的伪影抑制效果:高密度肌肉伪影ΔHU为29.0(9.0,41.0),AIX降幅达28.1(19.8,36.4);低密度肌肉伪影ΔHU为50.00(23.0,83.0),AIX降幅达34.1(7.5,60.7);(锁)骨伪影ΔHU为35.0(18.0,46.0),AIX降幅达45.6(22.2,69.0);血管伪影ΔHU(5.0(1.0,18.0))和甲状腺伪影ΔHU(14.0(7.8,27.5))的AIX值分别降幅达13.0(8.8,17.2)和14.7(12.3,17.1)。心脏起搏器周围伪影的ΔHU最大降幅达68.4%,AIX值降低61.5%。评估者间一致性良好(k=0.91,95%CI:0.88~0.93;k=0.88,95%CI:0.85~0.91;k=0.83,95%CI:0.79~0.86;k=0.81,95%CI:0.77~0.84;k=0.80,95%CI:0.76~0.83),iMAR组的主观评分显著高于非iMAR组。结论:单能级110 keV VMI联合iMAR算法可显著抑制心脏起搏器相关金属伪影。在抑制伪影与图像质量之间实现最佳平衡。该重建方案可作为PCCT评估起搏器患者的优选参数,具有明确临床应用价值。

       

      Abstract: Purpose: To evaluate the feasibility of combining photon-counting computed tomography (PCCT) virtual monoenergetic images (VMI) with iterative metal artifact reduction (iMAR) to suppress cardiac pacemaker-related metal artifacts and determine the optimal reconstruction energy level. Methods: This prospective study included patients with implanted cardiac pacemakers who underwent chest PCCT between February and December 2024; a total of 42 patients were enrolled in this study. Images were reconstructed at energy levels ranging from 70 to 150 keV (at 20-keV intervals), with and without iMAR. The artifact index (AIX) and corrected attenuation (ΔHU) were compared using the Friedman test. Interobserver agreement for subjective scoring was evaluated using weighted kappa analysis and subjective scores were compared using one-way repeated-measures ANOVA. Results: Compared with the non-iMAR group, the group subjected to the iMAR technique had significantly reduced AIX values across all artifact types, including the high/low-density muscle, clavicular, vascular, and thyroid regions. The 110 keV+iMAR reconstruction protocol demonstrated the optimal artifact suppression effect: for high-density muscle artifacts, the ΔHU was 29.0 (9.0, 41.0), with an AIX reduction of 28.1 (19.8, 36.4); for low-density muscle artifacts, the ΔHU was 50.0 (23.0, 83.0), with an AIX reduction of 34.1 (7.5, 60.7); for clavicular artifacts, the ΔHU was 35.0 (18.0, 46.0), with an AIX reduction of 45.6 (22.2, 69.0). Vascular and thyroid artifacts showed AIX reductions of 13.0 (8.8, 17.2) and 14.7 (12.3, 17.1), respectively, with corresponding ΔHU values of 5.0 (1.0, 18.0) and 14.0 (7.8, 27.5), respectively. The maximum reduction in ΔHU for artifacts surrounding the cardiac pacemaker reached 68.4%, with a 61.5% decrease in AIX. Interobserver agreement was excellent (kappa values ranging from 0.80 to 0.91), and the subjective scores in the iMAR group were significantly higher than those in the non-iMAR group. Conclusion: VMI at 110k eV combined with application of the iMAR algorithm significantly suppressed cardiac pacemaker-related metal artifacts and achieved an optimal balance between artifact reduction and image quality. This reconstruction protocol is recommended as the preferred parameter for evaluating patients with pacemakers using PCCT and has significant clinical value.

       

    /

    返回文章
    返回