Objective To explore the value of adaptive statistical iterative reconstruction (ASIR) and a sharp recon kernel to obtain high resolution pulmonary images in low-dose pediatric chest CT scans.Methods Totally 42 children underwent low-dose chest CT scans with ASIR were included.Age dependent noise index (NI) was used for dose optimization:NI=12 for 0-12 months old,NI=15 for ＞1 2 years old,NI=17 for 3-6 years old and NI=20 for ≥7 years old.Images were reconstructed to 0.625 mm using different recon kernels:Soft,Standard,Lung,and Chest kernel.ASIR blending was varied from 0 100％ to provide balanced image noise and spatial resolution.Two radiologists independently evaluated images for normal lung structures,abnormal CT findings and image noise on a 5 point scale with 3 being clinically acceptable.The best kernel,as well as the match with the best ASIR weight were analyzed statistically.Results CT images with lung kernel and ASIR 60％ were rated substantially better than those kernel.Conclusion ASIR 60％ with a sharp lung kernel can significantly improve image quality in low dose pediatric chest CT scans.%目的 探讨自适应迭代重建技术(ASIR)结合高分辨算法对儿童低剂量胸部CT肺部结构显示的影响.方法 回顾性分析接受低剂量胸部CT检查且存在肺内病变的患儿42例,0～12个月预设噪声指数12,＞1～2岁预设噪声指数15,3～6岁预设噪声指数17,≥7岁预设噪声指数20.将所有图像应用Soft、Standard、Lung、Chest分辨率模式重建为层厚0.625 mm的图像.以10％为步涨值,重建ASIR权重为0％～100％的11组图像.由2名医师分别用5分制评分法主观评价肺窗图像质量,包括图像主观噪声、正常肺结构及病变的显示能力,5分为最佳.统计学分析比较最佳的后处理算法,以及与之匹配的最佳ASIR权重.结果 Lung模式为观察肺部病变最佳的高分辨算法,ASIR 60％权重重建图像主观评分最佳.结论 采用ASIR 60％权重结合Lung高分辨算法可更好地显示儿童低剂量胸部CT的肺部结构.
展开▼
