Ono Shun, Niwa Tetsu, Yanagimachi Noriharu, Yamashita Tomohiro, Okazaki Takashi, Nomura Takakiyo, Hashimoto Jun, Imai Yutaka
Department of Radiology, Tokai University School of Medicine, 143, Shimokasuya, Isehara 259-1193, Japan.
Department of Radiology, Tokai University School of Medicine, 143, Shimokasuya, Isehara 259-1193, Japan.
J Neuroradiol. 2016 Feb;43(1):31-6. doi: 10.1016/j.neurad.2015.07.005. Epub 2015 Oct 28.
Iterative reconstruction (IR) offers noise reduction and improved image quality of computed tomography (CT). Our aim was to assess the imaging quality of non-contrast helical CT of the head in children using IR.
This study recruited 78 consecutive children aged ≤5 years (range: from 3 months to 5 years; mean: 1.7 years) who underwent an emergent non-enhanced helical CT of the head with no abnormal findings. The acquired data were reconstructed using filtered back projection (FBP) and sinogram-affirmed IR (SAFIRE) with strength levels of 2 (IR2) and 4 (IR4). The imaging quality of FBP, IR2 and IR4 was scored by two experienced neuroradiologists in terms of the contrast between the gray-white matter junction and artifacts from the skull at the level of the semioval center, basal ganglia and fourth ventricle. FBP, IR2 and IR4 scores were compared at each slice level. The signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated for FBP, IR2 and IR4 and were compared among the three reconstruction algorithms.
The score of IR2 and IR4 was significantly higher than that of FBP in terms of both the contrast between the gray-white matter junction and artifacts from the skull at each slice level (P<0.001). SNR and CNR on IR4 were the highest followed by those on IR2 and FBP (P<0.001).
IR may improve the image quality of helical CT of the head in children.
迭代重建（IR）可降低噪声并提高计算机断层扫描（CT）的图像质量。我们的目的是评估使用IR的儿童头部非增强螺旋CT的成像质量。
本研究纳入了78例年龄≤5岁（范围：3个月至5岁；平均：1.7岁）的连续儿童，他们接受了头部急诊非增强螺旋CT检查且无异常发现。采集的数据使用滤波反投影（FBP）和正弦图确认迭代重建（SAFIRE）进行重建，强度级别分别为2（IR2）和4（IR4）。两名经验丰富的神经放射科医生在半卵圆中心、基底节和第四脑室层面，根据灰白质交界与颅骨伪影之间的对比度，对FBP、IR2和IR4的成像质量进行评分。比较每个层面的FBP、IR2和IR4评分。计算FBP、IR2和IR4的信噪比（SNR）和对比噪声比（CNR），并在三种重建算法之间进行比较。
在每个层面上，IR2和IR4在灰白质交界与颅骨伪影之间的对比度方面的评分均显著高于FBP（P<0.001）。IR4上的SNR和CNR最高，其次是IR2和FBP（P<0.001）。
IR可能会提高儿童头部螺旋CT的图像质量。
