Department of Radiology, Vancouver General Hospital, Vancouver, BC, Canada.
Siemens Medical Solutions USA Inc., Mountain View, CA.
J Thorac Imaging. 2019 Nov;34(6):387-392. doi: 10.1097/RTI.0000000000000412.
The purpose of this study was to evaluate the clinical utility of temporal resolution optimization (TR-Opt), a computed tomography (CT) postprocessing technique, in reducing aortic motion artifacts in blunt thoracic trauma patients.
This was an IRB-approved study of 61 patients with blunt thoracic trauma carried out between February 18 and September 6, 2014; the patients had been imaged using a standardized dual-source high-pitch (DSHP) CT protocol. Image raw data were retrospectively postprocessed using the TR-Opt algorithm (DSHP-TR-Opt) and compared with conventional images (DSHP). Diagnostic ability to confidently identify and exclude potential injuries and qualitative aortic motion artifacts using a 5-point Likert scale (1=absence of motion artifacts; 5=severe motion artifact) was graded by 2 readers at multiple thoracic locations. Signal-to-noise and contrast-to-noise ratios were generated as quantitative indices of image quality.
Motion artifacts degrading interpretation and limiting diagnosis of aortic injuries were present in 45% (442/976) of the assessed regions on DSHP. TR-Opt algorithm eliminated motion artifacts in 85% of the motion-degraded areas (375/442), leaving persistent motion artifacts in only 15% (67/442). Motion artifacts were most improved at the interventricular septum (1±1 vs. 3±1), aortic valve (2±1 vs. 4±1.5), and ascending aorta (1±1 vs. 3±2, P<0.005). Mean aorta noise (NAo) was 41.7% higher in the DSHP-TR-Opt images (26.5 vs. 18.7 HU, P<0.0001).
Temporal resolution optimized reconstruction is a raw data-based CT postprocessing technique that can be used to remove the majority of thoracic aortic motion artifacts that commonly degrade interpretation when imaging blunt thoracic trauma patients.
本研究旨在评估时间分辨率优化（TR-Opt）的临床应用价值，该技术是一种计算机断层扫描（CT）后处理技术，可减少钝性胸部创伤患者的主动脉运动伪影。
这是一项在 2014 年 2 月 18 日至 9 月 6 日进行的、针对 61 例钝性胸部创伤患者的 IRB 批准的研究；这些患者使用标准的双源高（DSHP）CT 方案进行成像。使用 TR-Opt 算法（DSHP-TR-Opt）对图像原始数据进行回顾性后处理，并与常规图像（DSHP）进行比较。两位读者使用 5 分 Likert 量表（1=无运动伪影；5=严重运动伪影）对多个胸部位置的潜在损伤和主动脉运动伪影的识别和排除能力进行评分，以评估诊断能力。生成信号噪声比和对比噪声比作为图像质量的定量指标。
在评估的 976 个区域中，45%（442/976）的区域存在运动伪影，影响主动脉损伤的诊断。TR-Opt 算法消除了 85%（375/442）运动伪影区域的运动伪影，仅在 15%（67/442）的区域中仍存在运动伪影。运动伪影在室间隔（1±1 与 3±1）、主动脉瓣（2±1 与 4±1.5）和升主动脉（1±1 与 3±2）改善最明显，差异具有统计学意义（P<0.005）。DSHP-TR-Opt 图像的主动脉噪声（NAo）平均高 41.7%（26.5 与 18.7 HU，P<0.0001）。
时间分辨率优化重建是一种基于原始数据的 CT 后处理技术，可用于消除钝性胸部创伤患者成像时常见的、影响诊断的大多数主动脉运动伪影。
