双层探测器光谱CT运动补偿重组技术在改善冠状动脉CTA图像质量的应用  

Application of Motion Compensation Reconstruction in Dual-Layer Detector Spectral CT for Improving Coronary CTA Image Quality

作　　者：王伟平 吕铭亮 郑晓林 WANG Weiping;LV Mingliang;ZHENG Xiaolin(Department of Radiology,Dongguan Kanghua Hospital,Dongguan,Guangdong Province 523000,P.R China)

机构地区：[1]东莞康华医院放射科,东莞523000

出　　处：《临床放射学杂志》2025年第5期943-949,共7页Journal of Clinical Radiology

摘　　要：目的探讨双层探测器光谱CT运动补偿重组技术改善冠状动脉图像质量的效果。方法回顾性搜集行双层探测器光谱CT检查的119例疑似或确诊冠心病患者的临床和影像资料。根据扫描时心率分为低心率组(72例),心率<80次/分;高心率组(47例),心率≥80次/分;所有图像分别采用标准(STD)算法和运动补偿重组(MCR)算法。由两名放射科医师对STD算法和MCR算法的各组图像质量进行评分。测量主动脉根部(AO)、右冠状动脉中段(RCA)、左前降支中段(LAD)、左回旋支中段(LCX)CT值和噪声水平(SD),计算信噪比(SNR)和对比噪声比(CNR)。各组图像主、客观数据采用方差分析和卡方检验比较,使用Kappa检验分析两名放射科医师主观评价冠状动脉图像质量的一致性。结果比较低心率组、高心率组、所有病例组STD算法图像,低心率组各血管主观评分和优秀节段以及CT值、SNR、CNR显著高于高心率组和所有病例组,不可评估节段和噪声值明显低于高心率组和所有病例组,差异均有统计学意义(均P<0.05),可评估节段差异无统计学意义(P>0.05)。在MCR算法图像,可评估节段、各血管主观评分、优秀节段、不可评估节段、CT值、噪声值、SNR、CNR均优于STD算法,差异均有统计学意义(均P<0.05)。比较低心率组、高心率组、所有病例组MCR算法图像,可评估节段、各血管主观评分、优秀节段、不可评估节段、CT值、噪声值、SNR、CNR差异均无统计学意义(均P>0.05)。结论MCR算法能够减少运动伪影,提高冠状动脉主观图像评分和客观图像质量,特别是对高心率图像质量提升尤为显著,值得临床应用并推广。Objective To investigate the effectiveness of motion compensation reconstruction(MCR)technology in dual-layer detector spectral CT for improving coronary artery image quality.Methods Retrospective collection of clinical and imaging data from 119 patients with suspected or confirmed coronary heart disease who underwent dual-layer detector spectral CT examination in the radiology department of our hospital from April to November 2024.The patients were divided into two groups based on heart rate during scanning:low heart rate group(72 cases,HR<80 bpm)and high heart rate group(47 cases,HR≥80 bpm).All images were reconstructed using both standard(STD)and motion compensation reconstruction(MCR)algorithms.Two radiologists independently evaluated the image quality of STD and MCR reconstructions for all groups.Quantitative measurements included CT values and noise levels(standard deviation,SD)in the aortic root(AO),mid-right coronary artery(RCA),mid-left anterior descending artery(LAD),and mid-left circumflex artery(LCX).Signal-to-noise ratio(SNR)and contrast-to-noise ratio(CNR)were calculated.Subjective and objective data were compared using one-way ANOVA and chi-square tests.Inter-observer agreement for coronary image quality assessments was analyzed via Cohen’s Kappa coefficient.Results Compared with the STD algorithm images of the low heart rate group,high heart rate group,and all cases group,the low heart rate group had significantly higher subjective scores for each vessel,excellent segments,CT values,SNR,and CNR than the high heart rate group and all cases group,while the non-assessable segments and noise values were significantly lower than those of the high heart rate group and all cases group(all P<0.05).There was no significant difference in the assessable segments(P>0.05).In the MCR algorithm images,the assessable segments,subjective scores for each vessel,excellent segments,non-assessable segments,CT values,noise values,SNR,and CNR were all superior to those of the STD algorithm,with statistically signifi

关 键 词：冠状动脉CT血管成像 双层探测器光谱CT 心率 运动补偿重组技术 图像质量 

分 类 号：R816.2[医药卫生—放射医学]