摘要
目的:探讨新一代基于模型的迭代重建(MBIRn)中肺特异性设置(MBIR NR40和MBIR RP20)在低剂量胸部CT图像质量的临床应用价值。方法:前瞻性搜集胸部CT低剂量筛查患者30例,采用肺和标准算法ASIR40%、MBIRn中的肺特异性设置(MBIR NR40和MBIR RP20)3种算法重建图像,比较3组图像的客观质量(感兴趣区CT值、信噪比)以及MBIR NR40和MBIR RP20相对于标准算法ASIR40%的噪声降低率和SNR提高率。在肺算法ASIR40%、MBIR NR40和MBIR RP20评价肺内结构,在标准算法ASIR40%、MBIR NR40和MBIR RP20评价纵隔结构,由两名放射科医师采用双盲法阅片对图像质量进行主观评价。结果:与标准算法ASIR40%相比,MBIR NR40和MBIR RP20两种算法的图像肌肉噪声分别较ASIR40%降低80.78%、37.27%,脂肪噪声分别较ASIR40%降低69.27%、23.69%,差异具有统计学意义(P<0.05)。MBIR NR40和MBIR RP20两种算法图像肌肉SNR分别较ASIR40%组提高81.80%、37.64%,脂肪SNR分别较ASIR40%组提高72.98%、24.95%,差异具有统计学意义(P<0.05)。三种算法图像在主观噪声、肺内血管及支气管清晰度、纵隔内血管及淋巴结中评分均有统计学差异,其中MBIR NR40主观噪声最低,纵隔内结构显示最清晰,MBIR RP20肺内血管及支气管显示最清晰。结论:在低剂量胸部CT条件下,与ASIR40%相比,MBIR组能明显提高图像质量,其中MBIR RP20显示肺部结构最优。MBIR NR40可明显降低图像噪声,提高纵隔内结构显示清晰度。
Objective:To investigate the clinical value of lung specific settings(MBIR NR40 and MBIR RP20)in low-dose chest CT image quality in a new generation of model-based iterative reconstruction(MBIRn).Methods:The objective image quality the three groups(the value of CT in the region,the signal-to-noise ratio)and MBIR NR40 and MBIR RP20 relative to ASIR40%noise reduction rate and the increase rate of SNR were compared.And pulmonary structure was evaluated in lung algorithm of ASIR40%,MBIR NR40 and MBIR RP20,the mediastinal structure was evaluated in the standard algorithm of ASIR40%,MBIR NR40 and MBIR RP20.The subjective quality of the images was evaluated via double blind scanning method by two radiologists.Results:Compared with the standard algorithm ASIR40%,the image muscle noise of MBIR NR40 and MBIR RP20 decreased by 80.78%and 37.27%,and the image fat noise was 69.27%and 23.69%lower than ASIR40%,respectively.The difference was statistically significant(P<0.05).The SNR of MBIR NR40 and MBIR RP20 were increased by 81.80%and 37.64%,respectively,and the SNR of fat was increased by 72.98%and 24.95%,respectively;and the difference was statistically significant(P<0.05).Among them,MBIR NR40 has the lowest subjective noise,the most clear structures in the mediastinum and the most clear blood vessels and bronchi in the lungs of MBIR RP20.Conclusion:MBIR NR40 can significantly reduce image noise and improve the clarity of the structure display in the mediastinum.
作者
左秀娟
贺太平
张开元
贾永军
马光明
王咪
ZUO Xiu-juan;HE Tai-ping;ZHANG Kai-yuan(Department of Radiology,Honghui Hospital,Xi'an Jiaotong University,Shanxi 710054,China)
出处
《放射学实践》
北大核心
2018年第10期1082-1086,共5页
Radiologic Practice
关键词
体层摄影术
X线计算机
计算机模拟
统计计算
辐射剂量
Tomography,X-ray computed
Computer simulation
Statistical computing
Radiation dosag
