迭代算法在鼻窦多层CT中的应用

目的 探讨鼻窦多层CT（MSCT）检查中应用迭代算法（iDose算法）对图像质量及辐射剂量的影响。方法 选取53例鼻窦图像，计算其平均对比噪声比（CNR），以此值为基准选择合适的离体头颅标本作为研究对象。改变扫描参数（kV及有效mAs），分别使用滤波反投影算法（FBP）和迭代算法（iDose等级1~7）重建图像，对图像质量进行主、客观评价，找出可以符合诊断需求的参数组合。结果 冠状位图像的CNR高于水平位图像（t=-6.86，P<0.05）。120kV、100mAs/层、iDose5算法与常规扫描参数FBP算法所得图像CNR相当；100kV、60mAs/层、iDose4算法所得图像可满足对炎症的诊断，100kV、80mAs/层、iDose4算法所得图像可同时满足对炎症和外伤骨折的诊断，有效剂量可分别降低82%和75%。结论 使用iDose算法对鼻窦炎症及骨折的诊断，可以在保证满足诊断要求的前提下降低有效剂量。

Application of iterative algorithm in paranasal sinus MSCT

Objective To study the effect of iterative algorithm (iDose algorithm) on image quality and radiation dose in paranasal sinus MSCT.Methods A proper cadaveric head based on the average CNR (contrast noise ratio) of paranasal sinus images in 53 patients was applied.The cadaveric head was scanned with different parameters of kV and effective mAs.The images were reconstructed with filtered back projection and iterative algorithm (iDose 1-7) respectively.The subjective and objective evaluations were performed for all the images and the optimal parameter combination was found out.Results The CNR of coronal images was slightly higher than that of axial images (t=-6.86, P<0.05).The image CNR of protocol (120 kV,100 mAs/slice and iDose5) was close to that of conventional protocol.The protocol of 100 kV,60 mAs/slice and iDose4 offered the images adaptable for diagnosis of inflammation.The protocol of 100 kV, 80 mAs/slice and iDose4 offered the images adaptable for diagnosis of inflammation and fracture, and the radiation dose were reduced by 82% and 75%, respectively.Conclusions Using the iDose algorithm could help to offer diagnostic images for inflammation and fracture and reduce the radiation dose in the paranasal sinus MSCT.