Yamashiro Tsuneo, Miyara Tetsuhiro, Honda Osamu, Tomiyama Noriyuki, Ohno Yoshiharu, Noma Satoshi, Murayama Sadayuki
Department of Radiology, Graduate School of Medical Science, University of the Ryukyus, Nishihara, Okinawa, Japan.
Department of Radiology, Osaka University Graduate School of Medicine, Suita, Osaka, Japan.
Int J Chron Obstruct Pulmon Dis. 2015 Feb 11;10:321-7. doi: 10.2147/COPD.S74810. eCollection 2015.
To assess the advantages of iterative reconstruction for quantitative computed tomography (CT) analysis of pulmonary emphysema.
Twenty-two patients with pulmonary emphysema underwent chest CT imaging using identical scanners with three different tube currents: 240, 120, and 60 mA. Scan data were converted to CT images using Adaptive Iterative Dose Reduction using Three Dimensional Processing (AIDR3D) and a conventional filtered-back projection mode. Thus, six scans with and without AIDR3D were generated per patient. All other scanning and reconstruction settings were fixed. The percent low attenuation area (LAA%; < -950 Hounsfield units) and the lung density 15th percentile were automatically measured using a commercial workstation. Comparisons of LAA% and 15th percentile results between scans with and without using AIDR3D were made by Wilcoxon signed-rank tests. Associations between body weight and measurement errors among these scans were evaluated by Spearman rank correlation analysis.
Overall, scan series without AIDR3D had higher LAA% and lower 15th percentile values than those with AIDR3D at each tube current (P<0.0001). For scan series without AIDR3D, lower tube currents resulted in higher LAA% values and lower 15th percentiles. The extent of emphysema was significantly different between each pair among scans when not using AIDR3D (LAA%, P<0.0001; 15th percentile, P<0.01), but was not significantly different between each pair among scans when using AIDR3D. On scans without using AIDR3D, measurement errors between different tube current settings were significantly correlated with patients' body weights (P<0.05), whereas these errors between scans when using AIDR3D were insignificantly or minimally correlated with body weight.
The extent of emphysema was more consistent across different tube currents when CT scans were converted to CT images using AIDR3D than using a conventional filtered-back projection method.
评估迭代重建在肺气肿定量计算机断层扫描（CT）分析中的优势。
22例肺气肿患者使用相同的扫描仪，分别采用三种不同的管电流（240、120和60 mA）进行胸部CT成像。扫描数据使用三维自适应迭代剂量降低技术（AIDR3D）和传统的滤波反投影模式转换为CT图像。因此，每位患者生成了六次扫描图像，其中三次使用AIDR3D，三次不使用。所有其他扫描和重建设置均保持固定。使用商用工作站自动测量低衰减区百分比（LAA%；< -950亨氏单位）和肺密度第15百分位数。使用Wilcoxon符号秩检验对使用和不使用AIDR3D的扫描图像之间的LAA%和第15百分位数结果进行比较。通过Spearman秩相关分析评估这些扫描图像中体重与测量误差之间的关联。
总体而言，在每个管电流下，未使用AIDR3D的扫描系列的LAA%更高，第15百分位数更低（P<0.0001）。对于未使用AIDR3D的扫描系列，较低的管电流导致更高的LAA%值和更低的第15百分位数。在不使用AIDR3D时，各扫描图像之间肺气肿程度的差异均具有统计学意义（LAA%，P<0.0001；第15百分位数，P<0.01），但在使用AIDR3D时，各扫描图像之间的差异无统计学意义。在不使用AIDR3D的扫描中，不同管电流设置之间的测量误差与患者体重显著相关（P<0.05），而在使用AIDR3D的扫描中，这些误差与体重的相关性不显著或最小。
与传统的滤波反投影方法相比，当使用AIDR3D将CT扫描转换为CT图像时，不同管电流下肺气肿的程度更一致。
Zotero 插件