单能量去金属伪影算法重建效果的系统评价及Meta分析

Effects of single-energy metal artifact reduction algorithm: a systematic review and Meta-analysis

目的通过系统评价和Meta分析评估基于单能量去金属伪影算法的CT后处理技术(SEMAR)对金属伪影的去除效果。方法检索2009年1月至2018年1月美国生物医学数据库Pub Med、荷兰医学文摘Embase、Web of science、Cochrane图书馆和中国期刊全文数据库等发表的相关中英文文献,通过纳入及排除标准进行筛选,对纳入文献采用Cochrane风险偏移评估工具进行质量评估,提取文献中使用的评估标准及相关数据,并对相应数据进行合并。采用Review Manager 5.0对纳入文献汇总分析,根据文献异质性进行亚组分析及敏感性分析。结果符合纳入标准文献9篇,其中英文8篇,中文1篇。Meta分析显示,SEMAR技术可明显提升金属伪影和周围结构的评分,对金属伪影CT值、周围软组织噪声(SD值)具有良好的校正作用,MD伪影评分=1.19(95%CI:0.99~1.39)、M D前列腺/子宫图像质量评分=1.87(95%CI:0.69~3.05)、M D膀胱图像质量评分=2.31(95%CI:1.28~3.34)、M D伪影CT值=118.7HU(95%CI:50.48~186.93)、M D周围软组织噪声=298.05(95%CI:61.66~534.43);明显提高周围病变的检出率RD=0.07(0.02~0.13),但对周围组织的CT值[M D=-21.35(95%CI:-44.0~1.31)]及伪影指数[M D=3.11(95%CI:-0.97~7.15)]的影响无统计学意义。结论 SEMAR具有良好的去除金属伪影及提升周围病变检出效果,且在去除金属伪影的同时,对周围软组织的CT值无明显影响。

Objective To evaluate the effects of single-energy metal artifact reduction algorithm（SEM AR） by systematic reviewand M eta-analysis. Methods Studies on SEM AR and non-SEM AR technique published during Jan. 2009 to Jan. 2018 were searched in Pub M ed,Embase,Web of science,the Cochrane Library and Chinese National Knowledge Infrastructure（ CNKI）. Articles were reviewed and screened according to the inclusion and exclusion criteria. The quality was evaluated with Cochrane collaborations tool for assessing risk of bias. Data were extracted and then analyzed with ReviewM anager 5. 0 software. Subgroup analysis and sensitivity analysis were performed according to the heterogeneity of the literature. Results A total of 9 articles were enrolled,including 8 English articles and 1 Chinese article. M eta-analysis indicated that SEM AR could significantly improve the scores of the metal artifacts and surrounding organs,and correct the CT value and image noise： MD（ metal artifact score） = 1. 19（ 95% CI： 0. 99-1. 39）,MD（prostate/uterus image quality score） = 1. 87（95% CI： 0. 69-3. 05）,MD（bladder image quality score） = 2. 31（95%CI： 1. 28-3. 340）,MD（ metal artifact CT） = 118. 7 Hu（95% CI： 50. 48-186. 93）,MD（ surrounding soft tissue noise）= 298. 05（95% CI： 61. 66-534. 43）. It could also increase the detection rate of nearby lesions RD = 0. 07（0. 02-0. 13）,but did not affect the CT values of nearby structures[MD =-21. 35（ 95% CI：-44. 0-1. 31）] and artifact index[MD = 3. 11（95% CI：-0. 97-7. 15）]. Conclusion SEMAR has a good effect on the removal of metal artifacts and the detection of peripheral lesions,without affecting the CT values of surrounding soft tissues.