自适应统计迭代重建算法在头部低剂量CT扫描中的应用.pdfVIP

文章编号 １００６ ８１４７ （ ２０15 ） ０2 0161 03 作者简介张羽（1976 ）， 男， 技师， 研究方向： 放射学；E mail ： 1127296380@。 自适应统计迭代重建算法在头部低剂量 CT扫描中的 应用 张 羽 （天津市人民医院影像学部，天津 300121） 摘要 目的：使用自适应统计迭代重建技术（ASiR）算法，在宝石 CT进行头颅扫描，探讨头部低剂量扫描的可行性。 方法：患者 90例行头部 CT检查，随机分为 A、B、C 3组（每组 30例）。 3组病例扫描参数相同，A组单纯行滤波反投影算法（FBP）重建；B组 行 FBP重建后，打开减少剂量因子 DR，设定为 50％；C组行 ASiR重建后，打开减少剂量因子 DR，同样设定为 50％，其余扫描条 件 3组相同。对 3组图像分别测量空气及侧脑室前脚脑脊液 CT值的标准差，取均数作为图像噪声，由 3位医师对 3组数据进行 评分。 对 3组数据噪声值及评分分别进行方差分析和秩和检验。 结果：A组的容积 CT剂量指数（CTDI vol ）为 45.79 mGy；B组 为22.75 mGy；C组为 22.75 mGy，B、C组较 A组降低了 50.31%。 C组图像噪声（1.96±0.52）显著低于 B组（2.92±0.67）（P0.001）， 与 A组（２．１２±０．５３）无显著性差异（P=0.294）。 3组图像数据均能满足诊断要求，B组图像评分较低，A组和 C组无显著性差异。 结论：在颅脑 CT扫描中使用 ASiR算法，保证图像质量的同时可以大幅降低辐射剂量。 关键词 低剂量 CT扫描 ；自适应统计迭代重建技术；图像质量 中图分类号 R445 文献标志码 Ａ Application of adaptive statistical iterative reconstruction algorithm in low-dose head CT ZHANG Yu (Department of Radiology， Tianjin People’s Hospital, Tianjin 300121，China) Ａｂｓｔｒａｃｔ Oｂｊｅｃｔｉｖｅ： To explore the feasibility of low dose scan in head CT using adaptive statistical iterative reconstruction (ASiR) algorithm in gemstone CT.Methods: Ninety patients who received head CT scans were included, were randomly divided into three groups (n=30). Standard dose scan group and low dose scan group were treated with the same tube current 300 mA and 260 kV, respectively. After applying dose reduction factor (DR) 50% in B and C group, no differences were found in other scanning parameters. Standard dose scans were reconstructed with filter back projection (FBP) (group A) while low dose scans were performed with FBP (group B) and 50% ASiR (group C). For all 3 groups, s values of air and cerebrospinal fluid in anterior horns of lateral ventricles were measured, the mean value was taken as image noises, and all data were scored by 3 radiologists. Noise values and scores of 3 groups were compared by ANOVA and rank sum test. Results: CTDI vol was lowered by 50.31% in low dose group as compared to standard dose group. Noise measured in group C (1.96 ± 0.52) was