迭代算法在100kVp分次团注双期CTU检查中的应用研究  

The initial application of ASIR and MBIR in 100kVp split-bolus 2-phase CTU

作　　者：胡娟[1] 王鹤[1] 王霄英[1] 邱建星[1] 杨学东[1] 郭小超[1] 刘婧[1] 曹会志 

机构地区：[1]北京大学第一医院医学影像科,北京100034 [2]通用电气医疗集团CT影像研究中心,北京110006

出　　处：《放射学实践》2013年第5期509-513,共5页Radiologic Practice

摘　　要：目的:探讨不同比例自适应统计迭代重建(ASIR)、基于模型的迭代重建(MBIR)在100kVp分次团注双期CTU检查中的应用价值。方法:30例血尿待查患者分为两组,100kVp组15例,行全泌尿系平扫及分两次团注对比剂后行肾实质-肾盂期CT扫描;120kVp组15例,行全泌尿系CT平扫及单次团注对比剂后行实质期和肾盂期双期增强扫描。100kVp组的肾实质-肾盂期图像在100kVp条件下采集,其他所有期相的图像均在120kVp条件下采集。重建图像包括:100kVp组肾实质-肾盂期FBP、50%ASIR、100%ASIR、MBIR四组图像;120kVp组肾盂期FBP图像。两名阅片者对100kVp组和120kVp组CTU重建图像进行独立阅片,对图像质量进行主观评分,测量图像背景噪声、集合系统各段CT值,计算对比噪声比(CNR)。记录所有病例有效剂量(ED)和体型特异性剂量评估值(SSDE),并进行统计学分析。结果:100kVp组MBIR、50%ASIR与120kVp组图像质量主观评分差异无统计学意义(P>0.05),100kVp组FBP、100%ASIR图像质量评分低于120kVp组(P<0.05);100kVp组MBIR、100%ASIR图像噪声低于120kVp组(P<0.05),100kVp组50%ASIR与120kVp组图像噪声差异无统计学意义(P>0.05),100kVp组FBP图像噪声高于120kVp组(P<0.05);100kVp组MBIR图像各段尿路CNR高于120kVp组(P<0.05),100kVp组50%ASIR、100%ASIR与120kVp组图像各段尿路CNR差异无统计学意义(P>0.05),100kVp组FBP图像各段尿路CNR低于120kVp组(P<0.05)。100kVp组ED及SSDE均低于120kVp组(P<0.05)。结论:ASIR及MBIR技术可提高图像质量、增加图像对比,使常规剂量对比剂100kVp分次团注双期CTU达到或超过常规管电压单次团注三期CTU成像效果,从而明显降低X线辐射剂量。Objective：To evaluate the effect of adaptive statistical iterative reconstruction （ASIR） and model-based iterative reconstruction （MBIR） in 100kVp split-bolus 2 phase CTU. Methods： Thirty patients with hematuria recruited in this study were divided into two groups. 100kVp group： fifteen of them underwent split bolus 2-phase CTU examination. 120kVp group：the other fifteen patients underwent single-bolus 3-phase CT examination. The images of nephrographic-pyelographic phase of experimental group was acquired with 100kVp, and the reconstructed images included FBP, 50 % ASIR, 100% ASIR and MBIR. All the other phases of two groups were acquired with 120kVp,and the reconstruction method was FBP. Scores were assigned by two qualified readers to evaluate the image quality of each technique. Quantitative evaluation （noise,CNR） of each patient was measured segmentally. Results： There was no significant difference in the image quality between MBIR and 50% ASIR images of 100kVp group and 120kVp group （P=0. 095,0. 081）. The image quality scores of FBP and 100% ASIR images were lower than 120kVp group （P=0. 000,0. 004）. The noise of MBIR and 100% ASIR images were lower than 120kVp group （P=0. 000,0. 000）. There was no significant difference in the noise between 50% ASIR images and 120kVp group （P=0. 290）. The noise of FBP images of 100kVp group was significantly higher than that of 120kVp group （P= 0. 000）. The CNR of each segment of urinary tract on MBIR images of 100kVp group was significantly higher than that of 120kVp group （P=0. 000～0. 049）. There was no significant difference in CNR of urinary tract between 50% ASIR and 100% ASIR images of 100kVp group and 120kVp group （P=0. 059～0. 548）. The CNR of each segment of urinary tract on FBP images of 100kVp group was significantly lower than that of 120kVp group （P=0. 000～0. 029）. The ED and SSDE of 100kVp group were significantly lower than that of 120kVp group （P=0. 000,0. 003）. Conclusion： ASIR and MBIR can p

关 键 词：体层摄影术 X线计算机 辐射量 图像处理 计算机辅助 

分 类 号：R197.39[医药卫生—卫生事业管理] TP391.41[医药卫生—公共卫生与预防医学]