16层螺旋CT冠状动脉成像图像质量相关影响因素分析及最佳扫描条件选择(心脏体模模拟实验)

Analysis of Main Influence Factors on Image Quality of 16-slice Spiral CT Coronary Angiography and Optimization of Scanning Parameters (Cardiac Phantom Study)

目的采用新型心脏动态体模，对16层螺旋CT冠状动脉成像图像质量相关影响因素进行分析，并探讨在不同心率下的扫描速度和重组算法选择。材料与方法采用GE公司研制的新型心脏动态体模，心率设置为40-95次／min（间隔5次／min）共12组。采用GE LightSpeed 16层螺旋CT，以心脏冠状动脉扫描模式对不同心率下的心脏体模进行扫描，X线球管转速（即扫描速度）设置为0．5s／r和0．6s／r。所有的扫描数据在R-R间期90％时相分别以单扇区（Snapshot segment）、双扇区（Burst-2）和四扇区重组（Burst-4）3种心脏重组算法进行重组。所有重组数据传至AW4．1工作站行后处理成像。后处理方法采用容积再现（VR）、多平面重组（MPR）。分别对不同重组图像进行评分。统计学处理采用析因分析和多元线性回归分析。结果（1）总模型具有统计学意义（F=11．15，P〈0．0001）。不同心率（F=11．99，P〈0．0001）、不同转速（F=5．76，P=0．00196）、不同重组算法（F=9．21，P=0．0003）对图像质量有显著的影响，三者的交互作用均无显著差别（P〉0．05）；（2）不同心率间比较，心率≤65次／min时图像质量评分较高（P〈0．05）；（3）不同扫描速度比较，0．5s／r时，图像质量评分较高（P〈0．05）；（4）不同重组算法之间比较：重组算法为Burst-4和Burst-2时，图像质量评分较高（P〈0．05），但两者之间差异无统计学意义（P〉0．05）；（5）心率、扫描速度及重组算法与图像质量间具有多元线性回归关系（标准化回归系数分别为-0．70794、-0．16449和0．27341，F=34．43，P〈0．0001），各回归系数具有显著性。结论心率、扫描速度及重组算法等是影响16层螺旋CT冠状动脉成像图像质量的主要因素，可采用新型心脏动态体模进行评估。合理利用扫描参数能有效提高图像质量和检查的成功率。

Objective To discuss the main influence factors on image quality of 16-slice spiral CT coronary angiography and optimization of scanning time and reconstruction algorithms. Materials and Methods An adjustable pulsating cardiac phantom （GE） containing predetermined simulated coronary arteries was scanned using 16 slice helical CT （GE Light- Speed） with cardiac pulsating rates of 40,45,50,55,60,65,70,75,80,85,90, and 95 beats per minutes （bpm）. The rotation speed of 0.5 s and 0.6 s were used respectively. The raw data were reconstructed using one sector, two sector and four sector reconstruction algorithm at optimal window of the R-R interval. According to same evaluation standard of reformatted image, the image quality was scored by a radiologist and a technician. The correlation between heart rate （HR）, rotation Speed, reconstruction algorithm, and the score of image quality （SIQ） were analyzed. Factorial analysis was applied. The SIQ as independent variable and the HR, rotation speed, reconstruction algorithm as dependent variables were analyzed by multiple linear regression analysis. Results The heart rate, rotation speed and reconstruction algorithm had significant influence on the SIQ （ F = 11.15 ,P 〈 0. 0001 ）. The SIQ was highest when the heart rate was lower than 65 beats per minute; same situation was happened when rotation speed was 0.5 seconds per rotation. There was linear regression relationship between heart rate, rotation speed, reconstruction algorithm and the SIQ （P 〈 0.0001 ）. The multisector （ namely two sector and four sector ） reconstruction algorithm improved the image quality than one sector did. Conclusion The heart rate, rotation speed and reconstruction algorithm are the main influence factors on the image quality of coronary arteries, which could be evaluated with 16 slice helical CT using a pulsating cardiac phantom. Optimization of scanning parameters could improve the image quality of coronary arteries.