多层螺旋CT门静脉造影术准直与螺距匹配及重建方法优化选择

Optimal choice of collimation, pitch and reconstructive method in multi slice spiral CT portography

目的 探讨多层螺旋CT门静脉造影术 (MSCTP)准直与螺距匹配及重建方法的优化选择。方法 15 0例疑肝脏疾病患者 ,按不同螺距及准直随机分为 4组行MSCTP检查 ,对比分析不同组间及不同重建方法 (MIP、VR、SSD)间的图像质量。结果 不同准直间、不同螺距间CTP图像质量存在明显差异。一定扫描时间内 ,2 .5mm准直、0 .875螺距组CTP图像质量优于 1mm、1.75螺距组。不同重建方法间存在明显差异 ,以VR图像质量最好。结论 相同螺距、相同扫描条件下 ,准直越薄 ,三维图像质量越好。三种重建方法以VR显示门静脉图像质量最佳 ;SSD在显示立体结构空间定位准确性方面最佳。

Objective To investigate the optimal scan parameter (collimation and pitch) and the reconstructive methan pan pan pan pan pan pan pan ptography (MSCTP). Methods MSCTP was performed in 150 cases suspected with liver disease, which were randomly divided into 4 groups according to different collimation and pitch. VR, MIP and SSD were used to reconstruct the portal veins in all patients. All images were evaluated and scored. Results In the same pitch and reconstructive method, the difference of CTP quality between collimation had significant statistical difference. The thinner the collimation, the better the MSCTP quality. In the same conditions of collimation and reconstructive method, the difference of MSCTP quality between pitch had significant statistical difference, the group of pitch=0.875 was superior to the group of pitch=1.75. In the elementarily same scan time, the quality of CTP that the group of 2.5 mm collimation and pitch=0.875 was superior to the group of 1 mm collimation and pitch=1.75. In the same group, the difference between reconstructive methods had significant statistical difference, VR was the best method. Conclusion In the same pitch and the same scan conditions, the collimation was thinner, the CTP quality was better. In the four slice spiral CT scanning, during the same scan period (15 s), the quality of MSCTP that the group of 2.5 mm collimation and pitch=0.875 was superior to the group of 1 mm collimation and pitch=1.75. In the three reconstructive methods, VR was the best method in demonstrating portal vein, and SSD was the best method in the accuracy of space locating.