Riffel Philipp, Zoellner Frank G, Budjan Johannes, Grimm Robert, Block Tobias K, Schoenberg Stefan O, Hausmann Daniel
From the *Institute of Clinical Radiology and Nuclear Medicine, University Medical Center Mannheim, and †Computer-Assisted Clinical Medicine, Medical Faculty Mannheim, Heidelberg University, Mannheim; ‡Siemens Healthcare GmbH, Erlangen, Germany; and §Center for Advanced Imaging Innovation and Research (CAI2R), Department of Radiology, New York University School of Medicine, New York, NY.
Invest Radiol. 2016 Nov;51(11):714-719. doi: 10.1097/RLI.0000000000000299.
The purpose of the present study was to evaluate a recently introduced technique for free-breathing dynamic contrast-enhanced renal magnetic resonance imaging (MRI) applying a combination of radial k-space sampling, parallel imaging, and compressed sensing. The technique allows retrospective reconstruction of 2 motion-suppressed sets of images from the same acquisition: one with lower temporal resolution but improved image quality for subjective image analysis, and one with high temporal resolution for quantitative perfusion analysis.
In this study, 25 patients underwent a kidney examination, including a prototypical fat-suppressed, golden-angle radial stack-of-stars T1-weighted 3-dimensional spoiled gradient-echo examination (GRASP) performed after contrast agent administration during free breathing. Images were reconstructed at temporal resolutions of 55 spokes per frame (6.2 seconds) and 13 spokes per frame (1.5 seconds). The GRASP images were evaluated by 2 blinded radiologists. First, the reconstructions with low temporal resolution underwent subjective image analysis: the radiologists assessed the best arterial phase and the best renal phase and rated image quality score for each patient on a 5-point Likert-type scale.In addition, the diagnostic confidence was rated according to a 3-point Likert-type scale. Similarly, respiratory motion artifacts and streak artifacts were rated according to a 3-point Likert-type scale.Then, the reconstructions with high temporal resolution were analyzed with a voxel-by-voxel deconvolution approach to determine the renal plasma flow, and the results were compared with values reported in previous literature.
Reader 1 and reader 2 rated the overall image quality score for the best arterial phase and the best renal phase with a median image quality score of 4 (good image quality) for both phases, respectively. A high diagnostic confidence (median score of 3) was observed. There were no respiratory motion artifacts in any of the patients. Streak artifacts were present in all of the patients, but did not compromise diagnostic image quality.The estimated renal plasma flow was slightly higher (295 ± 78 mL/100 mL per minute) than reported in previous MRI-based studies, but also closer to the physiologically expected value.
Dynamic, motion-suppressed contrast-enhanced renal MRI can be performed in high diagnostic quality during free breathing using a combination of golden-angle radial sampling, parallel imaging, and compressed sensing. Both morphologic and quantitative functional information can be acquired within a single acquisition.
本研究旨在评估一种最近引入的自由呼吸动态对比增强肾脏磁共振成像（MRI）技术，该技术采用了径向k空间采样、并行成像和压缩感知相结合的方法。该技术允许从同一采集数据中回顾性重建两组运动抑制图像：一组具有较低的时间分辨率但图像质量更高，用于主观图像分析；另一组具有高时间分辨率，用于定量灌注分析。
在本研究中，25例患者接受了肾脏检查，包括在自由呼吸期间注射造影剂后进行的典型脂肪抑制、金角径向星状堆叠T1加权三维扰相梯度回波检查（GRASP）。图像在每帧55个辐条（6.2秒）和每帧13个辐条（1.5秒）的时间分辨率下重建。由两名盲法放射科医生对GRASP图像进行评估。首先，对低时间分辨率的重建图像进行主观图像分析：放射科医生评估最佳动脉期和最佳肾期，并根据5点李克特量表对每位患者的图像质量评分。此外，根据3点李克特量表对诊断置信度进行评分。同样，根据3点李克特量表对呼吸运动伪影和条纹伪影进行评分。然后，采用逐体素去卷积方法对高时间分辨率的重建图像进行分析，以确定肾血浆流量，并将结果与先前文献报道的值进行比较。
读者1和读者2对最佳动脉期和最佳肾期的总体图像质量评分中值均为4（图像质量良好）。观察到较高的诊断置信度（中值评分为3）。所有患者均未出现呼吸运动伪影。所有患者均存在条纹伪影，但未影响诊断图像质量。估计的肾血浆流量略高于先前基于MRI的研究报道（295±78 mL/100 mL每分钟），但也更接近生理预期值。
使用金角径向采样、并行成像和压缩感知相结合的方法，在自由呼吸期间可以高质量地进行动态、运动抑制的对比增强肾脏MRI检查。形态学和定量功能信息均可在一次采集中获得。
