Koc Gonca, Courtier Jesse L, Phelps Andrew, Marcovici Peter A, MacKenzie John D
Department of Radiology and Biomedical Imaging, UCSF Benioff Children's Hospital, 505 Parnassus Ave., San Francisco, CA, 94143-0628, USA.
Pediatr Radiol. 2014 Jul;44(7):787-94. doi: 10.1007/s00247-014-2899-y. Epub 2014 Feb 16.
Computed tomography (CT) is extremely important in characterizing blood vessel anatomy and vascular lesions in children. Recent advances in CT reconstruction technology hold promise for improved image quality and also reductions in radiation dose. This report evaluates potential improvements in image quality for the depiction of small pediatric vessels with model-based iterative reconstruction (Veo™), a technique developed to improve image quality and reduce noise.
To evaluate Veo™ as an improved method when compared to adaptive statistical iterative reconstruction (ASIR™) for the depiction of small vessels on pediatric CT.
Seventeen patients (mean age: 3.4 years, range: 2 days to 10.0 years; 6 girls, 11 boys) underwent contrast-enhanced CT examinations of the chest and abdomen in this HIPAA compliant and institutional review board approved study. Raw data were reconstructed into separate image datasets using Veo™ and ASIR™ algorithms (GE Medical Systems, Milwaukee, WI). Four blinded radiologists subjectively evaluated image quality. The pulmonary, hepatic, splenic and renal arteries were evaluated for the length and number of branches depicted. Datasets were compared with parametric and non-parametric statistical tests.
Readers stated a preference for Veo™ over ASIR™ images when subjectively evaluating image quality criteria for vessel definition, image noise and resolution of small anatomical structures. The mean image noise in the aorta and fat was significantly less for Veo™ vs. ASIR™ reconstructed images. Quantitative measurements of mean vessel lengths and number of branches vessels delineated were significantly different for Veo™ and ASIR™ images. Veo™ consistently showed more of the vessel anatomy: longer vessel length and more branching vessels.
When compared to the more established adaptive statistical iterative reconstruction algorithm, model-based iterative reconstruction appears to produce superior images for depiction of small pediatric vessels on computed tomography.
计算机断层扫描（CT）在描述儿童血管解剖结构和血管病变方面极为重要。CT重建技术的最新进展有望提高图像质量并降低辐射剂量。本报告评估了基于模型的迭代重建（Veo™）技术在描绘小儿小血管方面图像质量的潜在改善，该技术旨在提高图像质量并减少噪声。
与自适应统计迭代重建（ASIR™）相比，评估Veo™作为一种改进方法在小儿CT上描绘小血管的效果。
在这项符合健康保险流通与责任法案（HIPAA）且经机构审查委员会批准的研究中，17例患者（平均年龄：3.4岁，范围：2天至10.0岁；6名女孩，11名男孩）接受了胸部和腹部的增强CT检查。使用Veo™和ASIR™算法（通用电气医疗系统公司，威斯康星州密尔沃基）将原始数据重建为单独的图像数据集。四名盲法放射科医生对图像质量进行主观评估。评估肺、肝、脾和肾动脉所描绘分支的长度和数量。使用参数和非参数统计检验对数据集进行比较。
在主观评估血管清晰度、图像噪声和小解剖结构分辨率的图像质量标准时，读者表示更喜欢Veo™图像而非ASIR™图像。与ASIR™重建图像相比，Veo™重建图像中主动脉和脂肪的平均图像噪声明显更低。Veo™和ASIR™图像在平均血管长度和所描绘分支血管数量的定量测量上存在显著差异。Veo™始终显示出更多的血管解剖结构：更长的血管长度和更多的分支血管。
与更成熟的自适应统计迭代重建算法相比，基于模型的迭代重建在计算机断层扫描上描绘小儿小血管时似乎能产生更优质的图像。
