利用同源双能CT(SF-DECT)提高儿童腹部图像质量的可行性研究

Feasibility of improving the quality of children′s abdominal images using split-filter dual-energy CT

目的评估在儿童腹部检查中,同源双能CT(SF-DECT)是否能在低剂量条件下提高图像质量。方法采用儿童体模进行初步研究,并于2020年6月至12月期间前瞻性招募20名4~6岁的儿童进行临床验证。分别采用常规单能CT(SECT)和SF-DECT扫描体模和儿童的的腹部区域,比较SF-DECT和SECT在各剂量下(1、2、3和4 mGy)CT值、图像噪声、对比噪声比(CNR)和图像主观评分。结果在体模研究中,3和4 mGy剂量条件下,与SECT相比,SF-DECT的图像噪声分别降低了18.9%和23.6%,肝脏和肾脏CNR分别最多增加了12.8%和31.9%(Z=3.00、5.17,P<0.001)。在儿童SF-DECT扫描中,随着扫描剂量的增加,图像噪声降低(Z=4.64,P<0.001),肝脏和肾脏CNR增加(Z=3.78、3.39,P<0.001)。在体模研究和患者研究中,SF-DECT扫描图像的主观评分均优于SECT(Z=1.96~3.80,P<0.05)。结论与SECT相比,SF-DECT能够提高儿童腹部扫描图像质量。该技术在优化儿童腹部CT扫描方面有一定前景,但需要进行深入的临床研究来验证结果。

Objective To evaluate the effects of split-filter dual-energy CT(SF-DECT)in improving image quality at low doses in the process of abdominal examinations for children.Methods A preliminary study was conducted using child phantoms.Furthermore,20 children aged 4-6 years were recruited prospectively for clinical validation from June 2020 to December 2020.Conventional single-energy CT(SECT)and SF-DECT were employed to scan the abdominal areas of the phantoms and children.Then,the CT values,image noise,contrast to noise ratios(CNRs),and image subjective scores of SF-DECT and SECT were compared under various doses(1,2,3,and 4 mGy).Results For the phantoms under doses of 3 and 4 mGy,SF-DECT decreased the image noise by 18.9%and 23.6%,respectively,and increased the liver and kidney CNRs(CNRliv and CNRkid)by 12.8%and 31.9%at most,respectively,compared to SECT(Z=3.00,5.17,P<0.001).For children,SF-DECT decreased image noise(Z=4.64,P<0.001)and increased CNRliv and CNRkid(Z=3.78,3.39,P<0.001).For both the phantoms and the children,the subjective scores of images scanned using the SF-DECT were higher than those scanned using the SECT(Z=1.96-3.80,P<0.05).Conclusions Compared with SECT,SF-DECT can improve the quality of children's abdominal images.This technique has a certain prospect of optimizing abdominal CT for children.However,it is necessary to conduct in-depth clinical research to verify the result.