目的：探讨儿童副鼻窦宝石能谱CT扫描时，自适应性统计迭代重建（Adaptive Statistical Iterative Reconstruction，ASIR）的最佳权重和低剂量扫描参数的搭配。方法应用宝石能谱CT对20例儿童副鼻窦进行扫描，扫描条件为100 kV、50 mA，使用10%～100%权重的自适应迭代重建技术对本扫描条件下图像进行重建并进行噪声值（图像标准差SD值）、对比度噪声比（CNR）评价。请3位影像学专家对上述图像进行5分值主观评价，优选出最佳图像质量的权重。然后以此权重为基础，筛选儿童副鼻窦患儿100例，分为5组，常规剂量组（A）20例采用（120 kV，90 mA）扫描，使用滤波反向投影算法（Filtered Back Projection，FBP）对图像重建做为对照组；低剂量4组（B1，B2，B3，B4）各20例采用（100 kV，85 mA；100 kV，50 mA；80 kV，85 mA；80 kV，50 mA）扫描，对原始投影数据采用ASIR算法进行重建，其余条件不变。记录5组患者CT容积剂量指数（CT Dose Index of volume，CTDIvol）及剂量长度乘积（DLP）。比较5组图像的辐射剂量、图像质量及图像噪声、CT Dose Index of volume，CNR、并进行统计学分析。结果100 kV、50 mA扫描条件下图像的最佳ASIR权重为60%，低剂量组60%ASIR与常规剂量组FBP辐射剂量比较明显减低，B1，B2，B3，B4较A组CTDIvol分别降低了38.92%，63.81%，66.01%，80.04%；DLP分别降低了39.78%，64.83%，66.25%，80.23%；图像质量评分，B1，B2，B3较A组主观评分差异无统计学意义（P＞0.05）均能满足临床诊断要求，B4（4.34±0.64）较A组（4.57±0.22）明显低，B4噪声明显升高，图像不能满足临床诊断要求。结论儿童副鼻窦宝石能谱CT扫描时，ASIR算法最佳权重比例为60%，最适宜低剂量参数为B3（80 kV，85 mA）。%Objective To investigate the optimal ASIR (Adaptive Statistical Iterative Reconstruction) weight and low-dose scanning parameters in gem energy spectrum CT scanning of paranasal sinuses in children. Methods Gem energy spectrum CT was applied to perform paranasal scans in 20 cases of children. The scanning conditions were 100 kV, 50 mA, image reconstruction by using 10% to 100% weight of the adaptive iterative reconstruction technique under the present scanning conditions as well as the noise value (image standard deviation SD value) and CNR (Contrast-to-Noise Ratio) evaluation. Three radiologists were invited to make a subjective evaluation (ranging from 5 to 0 score) of the image and select the optimum image quality weight. And on the basis of optimum image quality weight, 100 cases of children with paranasal sinuses were screened and divided into five groups. 20 cases in Conventional Dose Group (Group A) were scanned under the conditions of (120 kV, 90 mA) and underwent image reconstruction by using FBP (Filtered Back Projection) algorithm as Control Group; 20 patients in four Low-Dose Groups (namely B1, B2, B3, B4) were scanned under the conditions of (100 kV, 85 mA; 100 kV, 50 mA; 80 kV,85 mA; 80 kV,50 mA) scan and underwent image reconstruction by using original projection data and ASIR algorithm; the rest conditions remained unchanged. The CT dose index volume (CTDIvol) and dose length product (DLP) of 5 groups were recorded and radiation dose, image quality, image noise and CNR of five groups were compared and statistically analyzed. Results The optimal ASIR weight was 60% under scanning conditions of 100 kV, 50 mA. Radiation dose of Low-Dose Group by using FBP at 60% ASIR was significantly lower than that of Conventional Dose Group; compared with Group A, CTDIvol and DLP of Group B1, B2, B3 and B4 decreased by 38.92%, 63.81%, 66.01%, 80.04% and 39.78%, 64.83% ,66.25%, 80.23%; there was no significant difference between the image quality ratings of compared with Group B1, B2, B3 and Group A (P>0.05), all of which met the requirements of clinical diagnosis. Compared with Group A (4.57±0.22), the image quality rating of Group B4 (4.34±0.64) was significantly lower with significantly increased noise, which could not meet the clinical diagnostic requirements. Conclusion For children with paranasal gems scanned by energy spectra CT, the optimal adaptive statistical iterative reconstruction weight was 60% and the most appropriate low dose parameters were B3 (80 kV, 85 mA).
展开▼
