中文摘要：

目的：探讨采用不同噪声指数（NI）联合迭代重组算法对不同体质量指数（BMI）患者行肝脏CT增强扫描的图像质量和辐射剂量的影响。方法前瞻性收集拟行肝脏CT增强扫描的160例患者纳入研究，根据BMI将患者分为A组（18 kg/m2≤BMI＜24 kg/m2）和B组（24 kg/m2≤BMI＜31 kg/m2），每组各80例，再将A、B组患者按照随机数字法各分为4个亚组，每个亚组20例。160例患者均行肝脏三期CT增强扫描，采用自动管电流调节技术，A、B两组的4个亚组NI分别设定为11、13、15、17，采用权重为50%的迭代重组算法进行重组。对各组图像图像质量进行主观评分和客观评价（信噪比和对比噪声比），并记录辐射剂量。采用方差分析或Kruskal-Wallis检验评价各组图像质量和辐射剂量的差异。结果 A组动脉期各亚组的图像质量评分差异有统计学意义（P＜0.05），其中NI=17组图像质量较其他3个亚组差。A组NI=17亚组动脉期图像的信噪比和对比噪声比与其他各组差异有统计学意义（P＜0.05），其余3组间的差异无统计学意义（P＞0.05）。A组NI=15与NI=17亚组的有效剂量（ED）差异无统计学意义（P＞0.05），其余各组两两比较差异均有统计学意义（P＜0.05）；NI=15与NI=17亚组的ED较NI=11组分别下降57.56%（2.17/3.77）和61.54%（2.32/3.77）。B组动脉期NI=15组的图像质量评分、信噪比和对比噪声比与其他各组差异有统计学意义（P＜0.05）；动脉期、门静脉期和延迟期NI=17组的图像质量评分、信噪比和对比噪声比与其他各组差异有统计学意义（P＜0.05）。B组NI=13、NI=15与NI=11亚组的ED差异有统计学意义（P＜0.05），NI=13与NI=15亚组的ED较NI=11组分别下降26.41%（1.69/6.40）和45.31%（2.90/6.40）。结论在保证影像质量前提下，根据BMI设置不同NI并联合应用迭代重组算法能够有效降低肝脏CT增强扫描的辐射剂量。

英文摘要：

Objective To investigate the effect of different noise index （NI） combined with iterative recombination on the image quality and radiation dose of CT scan in patients with different body mass index （BMI）. Method One hundred and sixty patients who had a liver CT enhanced scan were divided into group A （18 kg/m2≤BMI〈24 kg/m2 ） and group B （24 kg/m2≤BMI〈31 kg/m2） according to BMI, and each group had 80 patients. The two groups were randomly divided into 4 subgroups which NI value was 11, 13, 15 and 17 respectively. All images were restructured with 50% adaptive statistical iterative reconstruction. Subjective evaluation, objective evaluation [signal noise ratio （SNR） and contrast noise ratio （CNR）] and the＆amp;nbsp;effective dose of each group were recorded and calculated after the scan. ANOVA and Kruskal-Wallis test were used to evaluated the difference of imaging quality and radiation dose. Results In group A, the SNR, CNR and the subjective score of the later arterial phase images showed a statistical difference between NI=17 group and other subgroups （P〈0.05）, while there was no statistical differences among the other three subgroups during three CT enhanced phase. The average ED of NI=15, 17 group were decreased by 57.56%（2.17/3.77） and 61.54% （2.32/3.77） compared with NI=11 group, respectively. In group B, the SNR, CNR and the subjective image scores of the later arterial phase showed a significant difference between NI=15 and NI=11, 13 group （P〈0.05）. There was a statistically difference of image quality in group NI=17 compared with the other three subgroups in the later arterial phase, portal venous phase and equilibrium phase （P〈0.05）. The average ED of NI= 13, 15 group was decreased by 26.41% （1.69/6.40） and 45.31%（2.90/6.40） compared with NI=11 group, respectively. Conclusion Upon maintaining diagnostic imaging quality, setting different NI according to BMI and applying the iterative reconstruction algorithm can effectively reduce the rad