探究1024×1024重建矩阵结合Karl算法对CT门静脉成像的门脉图像质量和肝脏体积测量的影响

Impact of 1024×1024 Reconstruction Matrix Combined with Karl Algorithm on the Image Quality of Portal Vein and Liver Volume Measurement in CT Portal Venography

目的:探究1024×1024重建矩阵结合Karl算法对CT门静脉成像的门脉图像质量和肝脏体积测量的影响。方法:回顾性分析行全腹部增强CT检查的患者40例,所有患者均采用联影uCT760进行扫描。成像参数:管电压120 kV,管电流设置为剂量调制等级3,重建512×512矩阵,Karl 5级图像,获得图像为常规组(A组)。使用常规组的原始数据重建得到实验组(B组)图像:采用1024×1024矩阵,结合Karl 5、7、9级重建获得B_(1)~B_(3)组图像。应用联影智能科研平台系统测量肝脏体积,比较A、B两组门静脉和肝脏CT值、标准差(SD)值、对比噪声比(CNR)、肝脏体积值及图像主观评分。结果:B组门静脉和肝脏CT值与A组比较差异无统计学意义(P=0.079~0.766)。随着Karl等级的提高,B组门静脉和肝脏的SD值逐渐下降(P<0.001),CNR逐渐升高(P<0.001),肝脏体积值差异无统计学意义(P=0.999)。主观评价上,随着Karl等级的增加,B组伪影评分逐渐降低(P<0.05),图像噪声、门静脉边缘锐利度、远端血管显示清晰度及显示的门静脉分支数逐渐提高(P<0.05),B3组图像质量最佳,优于A组。结论:采用1024×1024重建矩阵结合Karl 9级重建算法能够优化门静脉图像质量,且准确测量肝脏体积。

Purpose:To explore the impact of reconstruction matrix of 1024×1024 combined with Karl algorithm on the image quality of portal vein and liver volume measurement in CT portal vein.Methods:A total of 40 patients who underwent contrast-enhanced abdominal CT scan were retrospectively analyzed.All patients were scanned by uCT760 with tube voltage of 120 kV and tube current modulation level 3.The images were reconstructed with a matrix of 512×512 and Karl algorithm level 5 for the conventional group(group A).The original data of the conventional group were reconstructed to obtain the images of the experimental group(group B).The images of groups B1 to B3 were reconstructed by a matrix of 1024×1024 combined with Karl algorithm level 5,7,and 9.The liver volume was measured by the United Imaging intelligent scientific research platform system.The CT value,standard deviation(SD)value,contrast-to-noise ratio(CNR)of portal vein and liver,liver volume and subjective image score were compared between groups A and B.Results:Compared with group A,there was no significant difference in the CT values of portal vein and liver in group B(P=0.079-0.766).With the increase of Karl grade,the SD values of portal vein and liver in group B decreased gradually(P<0.001),and the CNR increased gradually(P<0.001).Furthermore,no significant difference was observed in liver volumes within group B.With the increase of Karl grade,the artifact score of group B decreased gradually(P<0.05),while the noise score,the edge sharpness of portal vein,the clarity of distal vessels and the number of portal vein branches increased gradually(P<0.05).Group B3 was with the best image quality,which was better than that in group A.Conclusion:Using a reconstruction matrix of 1024×1024 combined with Karl 9 reconstruction algorithm can optimize the image quality of portal vein and measure the liver volume accurately.