1 Department of Radiology, Ansan Hospital, Korea University College of Medicine, 123 Jeokgeum-ro, Danwon-gu, Ansan-si, Gyeonggi, 15355, South Korea.
2 Medical Science Research Center, Ansan Hospital, Korea University College of Medicine, Ansan-si, South Korea.
AJR Am J Roentgenol. 2018 Sep;211(3):557-563. doi: 10.2214/AJR.17.19377. Epub 2018 Jul 5.
The purpose of this study is to determine the optimal energy level of virtual monochromatic images from spectral CT compared with conventional polychromatic images for reducing beam-hardening artifacts caused by contrast media in the thorax.
A total of 101 consecutive patients who underwent chest CT with contrast enhancement were retrospectively included in this study. The same contrast media and injection protocols were applied to the whole study population. Virtual monochromatic image datasets ranging from 70 to 200 keV and conventional polychromatic images were obtained. Readers' subjective image quality scores were recorded for conventional polychromatic and virtual monochromatic images obtained at 70, 100, 130, and 200 keV. Image noise, CT attenuation difference, contrast-to-noise ratio (CNR), and signal-to-noise ratio (SNR) were also obtained in each algorithm. Comparisons of parameters between algorithms were performed.
The best subjective image quality score and significantly lower image noise were observed for 130-keV virtual monochromatic images compared with conventional polychromatic images (all p < 0.001). Also, CT attenuation differences were significantly lower for both 100- and 130-keV virtual monochromatic images than for conventional polychromatic images (all p < 0.001). Meanwhile, the lowest differences in CT attenuation were observed for 100-keV virtual monochromatic images compared with conventional polychromatic images. However, there were no significant differences in CT attenuation between 100- and 130-keV virtual monochromatic images. SNR was similar between 130-keV virtual monochromatic images and conventional polychromatic images, although both SNR and CNR decreased as the energy level increased.
Virtual monochromatic imaging reduced beam-hardening artifacts and improved image quality, and optimal evaluation of chest CT was best achieved at 100 and 130 keV.
本研究旨在确定与传统多光谱图像相比，光谱 CT 的虚拟单能量图像的最佳能量水平，以减少胸部对比剂引起的硬化伪影。
回顾性纳入 101 例接受胸部 CT 增强检查的连续患者。对所有研究人群均应用相同的对比剂和注射方案。获得虚拟单能量图像数据集，范围从 70 keV 到 200keV，以及传统多光谱图像。记录常规多光谱图像和在 70keV、100keV、130keV 和 200keV 时获得的虚拟单能量图像的读者主观图像质量评分。还获得了每种算法的图像噪声、CT 衰减差异、对比噪声比（CNR）和信噪比（SNR）。比较了算法之间的参数。
与常规多光谱图像相比，130keV 虚拟单能量图像的最佳主观图像质量评分和明显更低的图像噪声（均 p<0.001）。此外，100keV 和 130keV 虚拟单能量图像的 CT 衰减差异明显低于常规多光谱图像（均 p<0.001）。同时，与常规多光谱图像相比，100keV 虚拟单能量图像的 CT 衰减差异最小。然而，100keV 和 130keV 虚拟单能量图像之间的 CT 衰减差异无统计学意义。130keV 虚拟单能量图像的 SNR 与常规多光谱图像相似，尽管随着能量水平的增加，SNR 和 CNR 均降低。
虚拟单能量成像减少了硬化伪影并改善了图像质量，最佳胸部 CT 评估在 100keV 和 130keV 时最佳。
