不同算法与肾强化程度对肾囊肿假性强化的影响

摘要

目的:采用试验模型研究不同重建算法和肾强化程度对肾囊肿假性强化的影响.方法:设计一个可变化浓度的模拟肾脏的体模和模拟不同直径肾内囊肿的试管作为研究对象.将肾浓度设定为40、100、140及240 HU.两种MSCT机均采用腹部螺旋扫描模式,层厚分别为5 mm和1 mm.Siemens双能CT使用B30和B70两种重建算法,GEMSCT机使用standard和bone两种重建算法.对比分析肾强化程度及重建算法对肾囊肿CT值测定的影响；同时比较囊肿在不同层厚下CT值的差异,分析产生假性强化的原因.结果:假性强化在两种CT机型和不同的重建算法中均可发生,假性强化程度为11.47～17.46HU.虽然两种CT机均可出现假性强化现象,但GE MSCT机发生的概率高于Siemens双能CT.Siemens双源CT仅在240 HU背景浓度、B70算法时出现囊肿假性强化,重建算法对其影响较大.背景浓度为240 HU时GE MSCT机均可发生假性强化现象,GE MSCT机不同重建算法和层厚对是否产生肾囊肿假性强化影响不大(P＞0.05).结论:肾囊肿假性强化是客观存在的,它受多种因素的影响.重建算法的B值越大出现假性强化现象的概率越高,较高的肾脏强化程度同样会增加假性强化发生的概率.%Objective: To investigate the dependence of renal cyst pseudo-enhancement on multidetector computed to-mographic (CT) convolution kernel and renal attenuation in a phantom. Methods; Renal cyst pseudo-enhancement was examined by using a rectangular water plastic tank (30X21 X 18cm) that simulated the human body. A kidney phantom was created to accept interchangeable 40,100,140 and 240HU attenuation that contained pure water cylinders measuring 5,13 and 18mm in diameter. The GE Light Speed Pro 32 and Siemens Somatom Definition Flash CT were used in the study. Both B30 and B70 convolution kernels were selected in Siemens dual-source CT scanner, and both standard and bone convolution kernel in GE scanner. Reconstruction slice thickness was 5mm and lmm respectively. This study emphasized particularly on analyzing the influence of renal attenuation and algorithms to the attenuation of renal cysts. The renal cysts attenuation on multidetector CT scanner was analyzed and reconstruction thickness was measured at the same time. Results: Pseudoen-hancement (range,11. 47 - 17, 46HU) was observed by using both scanners and different reconstruction algorithm. Although pseudoenhancement was observed with both MDCT scanners,the effect of GE MDCT scanner was higher than that of dual-source CT. The dual-source CT only presented pseudo-enhancement on the "B70" in two "cysts" in the background attenuation 240HU. Pseudoenhancement is strongly dependent on dual-source CT convolution kernel. All of cysts presented "pseudoenhancement" phenomenon on 240HU attenuation GE MDCT scanner. Pseudoenhancement was significantly higher with a background renal density of 240HU (P<0. 05). The essential reason of pseudoenhancement on GE MDCT was not dependent on the reconstruction algorithm and slice thickness (P>0. 05). Conclusion; Varying the parameter may mitigate this phenomenon, which is independent of volume-averaging effects. Pseudo-enhancement more likely occurred in high-spatial-resolution kernels and high levels of renal enhancement.