基于模型的迭代重建算法优化低管电压儿童胸部CTA图像质量

Full model-based iterative reconstruction in optimizing image quality in low tube voltage chest CTA images of children

目的评价基于模型的迭代重建算法(MBIR)优化低管电压儿童胸部CTA图像质量的价值。方法收集接受胸部低管电压(80kVp)CTA的儿童患者41例,对原始数据进行MBIR、30%自适应迭代重建(ASiR)与70%滤波反射投影(FBP)混合及FBP重建,评价图像整体质量及细小血管显示能力,测量降主动脉、同层面背部肌肉和脂肪噪声值,计算降主动脉SNR和CNR。结果 MBIR图像整体质量评分和细小血管显示能力评分为均明显优于30%ASiR与70%FBP混合图像及FBP图像(P均<0.05);MBIR图像中各组织噪声值明显低于30%ASiR与70%FBP混合图像和FBP图像(P均<0.05),降主动脉SNR和CNR(28.30±7.85、24.61±7.52)明显高于30%ASiR与70%FBP混合图像(15.80±5.73、14.19±4.22)和FBP图像(12.06±3.66、10.59±3.53,P均<0.05)。结论 MBIR用于处理儿童低管电压胸部CTA图像,图像质量明显高于ASiR及FBP图像,值得临床推广应用。

Objective To evaluate the feasibility of full model-based iterative reconstruction（MBIR）in optimizing image quality in low tube voltage chest CTA images of children.Methods Totally 41 children who underwent low tube voltage（80kVp）chest CTA were included.The images were retrospectively reconstructed with MBIR,30% of adaptive statistical iterative reconstruction（ASiR）＋70% filtered back-projection（FBP）and FBP.The image quality and ability to display small arteries were evaluated,the image noise value of descending aorta,muscle and fat were measured and SNR,CNR of descending aorta were calculated.Results The score of image quality and ability to display small vessels were both higher in MBIR image compared with 30% ASiR＋70% FBP image and FBP image（all P〈0.05）.The noise value of each tissue in MBIR image was significantly lower than that in 30% ASiR＋70% FBP image and FBP image（all P〈0.05）.SNR and CNR of descending aorta in MBIR image（28.30±7.85,24.61±7.52）were both significantly higher compared with 30%ASiR＋70% FBP image（15.80±5.73,14.19±4.22）and FBP image（12.06±3.66,10.59±3.53,all P〈0.05）.Conclusion Compared with FBP and ASiR image,the image quality is significantly improved by the MBIR in children low tube voltage chest CTA,which is worthy of popularizing.