低浓度对比剂门静脉成像在能谱CT中的临床应用

目的：探讨低浓度对比剂门静脉成像在能谱CT中的临床应用。方法50例患者分为5组行肝脏门静脉期能谱CT扫描，患者注射对比剂总量为1 mL/kg，5组患者注射对比剂浓度分别为320 mgI/mL，288 mgI/mL，256 mgI/mL，224 mgI/mL，192 mgI/mL，扫描结束获得门静脉期0.625 mm层厚的门静脉最佳对比度噪声比（CNR）单能量图像，用此单能量图像对门静脉行VR和MIP重组。测量5组图像中门静脉和相同层面背部肌肉竖脊肌及腹壁脂肪的CT值及其标准差，计算信噪比（SNR）及CNR并取平均值，由2名高年资医师在盲法下对5组图像进行主观图像质量评分并取平均值，5组间比较采用单因素方差分析（ANOVA）。2名医师对重建图像质量一致性评估采用Kappa检验。结果门静脉最佳CNR单能量图像能量为50~55 keV。5组注射不同浓度对比剂患者的SRN、CNR分别为15.83±3.02、14.78±2.13、13.87±1.91、9.37±2.13、7.72±2.82和14.95±3.82、15.01±3.79、13.93±3.52、10.35±4.22、9.65±3.56。2位医师评分值均值为4.50±0.81、4.45±0.76、4.41±0.73、4.38±0.79、3.5±0.57。浓度为320 mgI/mL、288 mgI/mL、256 mgI/mL之间SNR、CNR及2位医师评分值之间差异无统计学意义（F=36.221，P〈0.05）；320 mgI/mL、288 mgI/mL、256 mgI/mL与224 mgI/mL、192 mgI/mL之间SNR、CNR及3位医师评分值之间差异有统计学意义（F=7.221，P〉0.05）。2名医师对320 mgI/mL、288 mgI/mL、256 mgI/mL对比剂浓度产生图像质量评价的一致性（Kappa值为0.86）高于224 mgI/mL、192 mgI/mL对比剂浓度产生图像（Kappa值为0.63）。结论应用能谱CT最佳单能量门静脉成像能显著降低对比剂使用浓度，推荐使用对比剂浓度为256 mgI/mL，注射总量为1 mL/kg。推荐能谱CT显示门静脉的最佳单能量为50～55 keV。

The application with low concentration of contrast media and spectral cT imaging in the portal venography

Objective To investigate the effect of low concentration of contrast media with energy spectral CT on the image quality of CT portal venography. Methods 50 patients were divided into 5 groups underwent spectral CT examination on a GE Discovery CT 750 HD scanner, the total contrast media were offered at 1 mL/kg to 50 patients, the contrast media concentration of 320 mgI/mL, 288 mgI/mL, 256 mgI/mL, 224 mgI/mL, 192 mgI/mL were offered for the 5 groups patients. During portal phase, 0.625 mm slice thickness polychromatie images and optimal monochromatic images were obtained, and volume rendering and maximum intensity projection were created to show the portal venography respectively, the CT unit and standard deviation were measured at the port venography and the erector muscle of spine, calculated the signal-noise ratio and the contrast-noise ratio, calculated the singal-noise ratio and contrast-noise ratio with the CT volue of port venography and the erector muscle of spine. The overall imaging quality was evaluated on a ifve-point scale by 3 radiologists, Comparisons between 5 groups were tested by One-Way ANOVA analysis. Results The monochromatic images at 50-55 keV were found to demonstrate the best CNR for both the portal vein and erector muscle of spine. The SNR with the different concentrations of contrast media was 15.83±3.02, 14.78±2.13, 13.87±1.91, 9.37±2.13, 7.72±2.82 and the CNR was 14.95±3.82, 15.01±3.79, 13.93±3.52, 10.35±4.22, 9.65±3.56. The values by three independent radiologists were 4.50±0.81, 4.45±0.76, 4.41±0.73, 4.38±0.79, 3.5±0.57. The dates of SNR , CNR and values were not statistically signiifcant at the contrast media concentration of 320 mgI/mL, 288 mgI/mL, 256 mgI/mL, 224 mgI/mL（F=36.221, P〈0.05）. The dates of SNR, CNR and values were statistically signiifcant at the contrast media concentration of 320 mgI/mL, 288 mgI/mL, 256 mgI/mL and 224 mgI/mL, 192 mgI/mL（F=7.221, P〉0.05）. In the comparison of image quality, Kappa value was 0.86 with the contrast media concen