综合评价高级建模迭代重建在颈部低剂量CT扫描中对图像质量的影响

A comprehensive assessment of the effects of advanced modeled iterative reconstruction on image quality in low-tube-voltage neck CT

目的评价高级建模迭代重建(ADMIRE)算法与传统滤波反投影(FBP)算法在颈部低剂量CT扫描中对图像质量的综合影响。方法收集80例需行颈部CT检查的病例进行低剂量CT扫描。所有扫描病例行CT图像重建时均采用四种不同的重建算法(FBP、ADMIRE 1、ADMIRE 3、ADMIRE 5)分别进行轴位重建,然后对获得的重建图像进行主、客观评价,最后得出图像CT值、图像噪声、SNR、CNR及整体图像质量等的对比情况。结果图像CT值方面各重建算法没有差别,图像噪声高低排序为ADMIRE 1>FBP>ADMIRE 3>ADMIRE 5,图像SNR与CNR高低排序为ADMIRE 5>ADMIRE 3>FBP>ADMIRE 1,图像质量主观评分优劣排序为ADMIRE 3>ADMIRE 5>FBP>ADMIRE 1。结论在颈部低剂量CT扫描中,相比传统的FBP算法,ADMIRE算法(ADMIRE1除外)不仅能大幅度降低图像的噪声,也能够显著提升图像的整体质量。

Objective To evaluate the comprehensive effects of advanced modeled iterative （ADMIRE） reconstruction and filtered back-projection （FBP） reconstruction on image quality in low-tube-voltage neck CT. Methods A total of 80 patients underwent low-tube-voltage neck CT examination using the second-generation dual-source CT. Image series were reconstructed with standard FBP and ADMIRE strength levels 1, 3, and 5 in axial reconstruction. Then, we subjectively and objectively evaluated the obtained image. The compared results of CT value, image noise, SNR, CNR and overall image quality can finally be obtained. Results There is no difference among the four algorithms in the CT value of the images. The image noise ranks in the following order ADMIRE 5〉ADMIRE 3〉FBP〉ADMIRE 1, while the order for SNR and CNR of images is ADMIRE 3〉ADMIRE 5〉FBP〉ADMIRE 1. The image quality is subjectively graded in the following order ADMIRE 3〉ADMIRE 5〉FBP〉ADMIRE 1. Conclusion Compared with the traditional algorithm of FBP, ADMIRE（ADMIRE1 not included） can not only reduce image noise but also enhance the overall image quality to a great extent in lose-dose CT scan.