目的：探讨基于体质量指数（BMI，kg/m2）的个性化噪声指数（Noise index，NI）设定在腹部CT成像中的应用价值。方法：采用固定NI（10 HU）的211例行腹部CT平扫患者作为A组，不同NI（10～14 HU，间隔2 HU）109例患者作为B组，A、B两组按BMI（BMI≤23、＞23~26、＞26）各分成3个亚组：A1（47例）、A2（93例）和A3（71例）；B1（42例）、B2（35例）和B3（32例）。所有病例均在GE Discovery HD CT上完成，A组扫描参数为120 kVp，3D自动mA和NI=10 HU，扫描时间0.8 s，层厚5 mm，重建间隔5 mm，螺距1.375∶1，图像处理40% ASIR，重建方式Stnd。B组采用个体化扫描方案：B1组NI=10 HU、B2组NI=12 HU、B3组NI=14 HU，余扫描参数同A组。测量肝脏标准差（SD），记录容积剂量指数（CTDIvol）；A、B两组的组内及组间的SD、CTDIvol比较采用独立样本t检验；对图像质量进行５分制主观评价，评价结果的一致性检验采用Kappa分析。结果：SD值：A1、A2、A3分别为（11.01±0.88） HU、（9.44±1.24） HU和（8.30±1.40） HU；B1、B2、B3分别是（10.90±0.82） HU、（10.79±0.52） HU和（10.96±0.68） HU；A1>A2>A3有统计学差异（P<0.05）；B组内SD值无统计学差异（P>0.05）；A1与B1无统计学差异（P>0.05），A2<B2、A3<B3差异有统计学意义（P<0.05）。CTDIvol：A1、A2、A3分别为（7.82±1.48） mGy、（15.00±4.21） mGy和（23.42±5.31） mGy；B1、B2、B3分别为（7.33±1.49） mGy、（8.45±2.35） mGy和（9.54±2.84） mGy；A1与B1无统计学差异，B2比A2降低43.67%，B3比A3降低59.26%；图像评分：A1、A2、A3分别为4.57±0.62、4.58±0.61、4.59±0.62；B1、B2、B3分别为4.57±0.50、4.80±0.65、4.56±0.67。A、B组内及组间图像质量对比无统计学差异（P>0.05）；2名医师对图像质量评分的Kappa一致性检验结果分别为0.835、0.711、0.734、0.742、0.809、0.761。结论：使用自动管电流调制技术（ATCM）行腹部CT扫描，根据病人的BMI调整NI值的设置，较大BMI病人的辐射剂量有较大幅度降低，不影响图像诊断质量。

Objective: To explore the value of personalized noise index(NI) setting based on body mass index(BMI) in abdominal CT imaging. Methods: Group A contained 211 patients who underwent abdominal ATCM-CT at a fixed NI(10 HU), while group B contained 109 patients who underwent abdominal ATCM-CT at different NIs(10~14 HU, interval: 2 HU). Either group A or B was further divided into three subgroups based on BMI(BMI≤23, ＞23~26, >26): A1(n=47), A2(n=93) and A3(n=71); B1(n=42), B2(n=35) and B3(n=32). The examinations of all the cases were completed on GE Discovery HD CT. The following scan parameters were used in group A: 120 kVp, 3D Auto mA, NI=10 HU, scan time=0.8 s, slice thickness=5 mm, reconstruction interval=5 mm, pitch=1.375∶1, image processing mode: 40% ASIR, and reconstruction mode: Stnd. The individualized scan plan was used in group B: B1: NI=10 HU, B2: NI=12 HU, B3: NI=14 HU; the remaining parameters were the same as group A. Three levels adjacent to hepatic hilum were selected for each group of images. The SD of liver with a uniform density was measured and the results were averaged as the final image SD, and the volume-dose index(CTDIvol) was recorded. The inter-group and intra-group comparisons of SD and CTDIvol of groups A and B were performed with t test. The image quality was subjectively evaluated by two senior radiologists(working experience more than 5 years) with a 5-point scale, and the inter-observer consistency was analyzed by Kappa test. Results: The SD of subgroups A1, A2 and A3 was (11.01±0.88) HU, (9.44±1.24) HU and (8.30±1.40) HU, respectively, and that of subgroups B1, B2 and B3 was (10.90±0.82) HU, (10.79±0.52) HU and (10.96±0.68) HU, respectively. The results showed a significant difference within group A(A1>A2>A3)(P<0.05), but not within group B(P>0.05). There was no significant difference between subgroups A1 and B1(P>0.05), but A2<B2 and A3<B3 with significant differences(P<0.05). The CTDIvol of subgroups A1, A2 and A3 was (7.82±1.48) mGy, (15.00±4.21) mGy and (23.42±5.31) mGy, respectively, and that of subgroups B1, B2 and B3 was (7.33±1.49) mGy, (8.45±2.35) mGy and (9.54±2.84) mGy, respectively. The results showed no significant difference between subgroups A1 and B1, but B2<A2(a decrease by 43.67%) and B3<A3(a decrease by 59.26%) significantly. The image score of subgroups A1, A2 and A3 was 4.57±0.62, 4.58±0.61 and 4.59±0.62, respectively, and that of subgroups B1, B2 and B3 was 4.57±0.50, 4.80±0.65 and 4.56±0.67, respectively. The inter-group and intra-group comparisons of group A and B showed no significant difference in image quality(P>0.05). The results of Kappa test for the consistency of image quality score between two senior radiologists were 0.835, 0.711, 0.734, 0.742, 0.809 and 0.761. Conclusion: Using abdominal ATCM-CT based on BMI-dependent NI adjustment, there is a great reduction of radiation dose for patients with a bigger BMI, without influencing the diagnostic quality of images.