From the Departments of Radiology (C.J.M., S.F., J.R.P., N.P., M.A.G., D.M., C.A.F., Y.C.), Genetics and Genome Sciences (M.L.D.), Pediatrics (M.L.D., K.M.D., C.A.F.), and Biomedical Engineering (M.A.G., D.M., C.A.F.), Case Western Reserve University, 11100 Euclid Ave, Bowell Building, Room B131, Cleveland, OH 44106; Departments of Radiology (M.M.) and Pediatrics (K.M., K.K.), University Hospitals Cleveland Medical Center, Cleveland, Ohio; and Center for Pediatric Nephrology, Cleveland Clinic Children's Hospital, Cleveland, Ohio (A.P., K.M.D.).
Radiology. 2021 Aug;300(2):380-387. doi: 10.1148/radiol.2021202302. Epub 2021 Jun 8.
Background MR fingerprinting (MRF) provides rapid and simultaneous quantification of multiple tissue parameters in a single scan. Purpose To evaluate a rapid kidney MRF technique at 3.0 T in phantoms, healthy volunteers, and patients. Materials and Methods A 15-second kidney MRF acquisition was designed with 12 acquisition segments, a range of low flip angles (5°-12°), multiple magnetization preparation schema (T1, T2, and fat suppression), and an undersampled spiral trajectory. This technique was first validated in vitro using standardized T1 and T2 phantoms. Kidney T1 and T2 maps were then obtained for 10 healthy adult volunteers (mean age ± standard deviation, 35 years ± 13; six men) and three pediatric patients with autosomal recessive polycystic kidney disease (ARPKD) (mean age, 10 years ± 3; two boys) between August 2019 and October 2020 to evaluate the method in vivo. Results Results in nine phantoms showed good agreement with spin-echo-based T1 and T2 values ( > 0.99). In vivo MRF kidney T1 and T2 assessments in healthy adult volunteers (cortex: T1, 1362 msec ± 5; T2, 64 msec ± 5; medulla: T1, 1827 msec ± 94; T2, 69 msec ± 3) were consistent with values in the literature but with improved precision in comparison with prior MRF implementations. In vivo MRF-based kidney T1 and T2 values with and without B correction were in good agreement ( > 0.96, < .001), demonstrating limited sensitivity to B field inhomogeneities. Additional MRF reconstructions using the first nine segments of the MRF profiles (11-second acquisition time) were in good agreement with the reconstructions using 12 segments (15-second acquisition time) ( > 0.87, < .001). Repeat kidney MRF scans for the three patients with ARPKD on successive days also demonstrated good reproducibility (T1 and T2: <3% difference). Conclusion A kidney MR fingerprinting method provided in vivo kidney T1 and T2 maps at 3.0 T in a single breath hold with improved precision and no need for B correction. © RSNA, 2021 See also the editorial by Laustsen in this issue.
背景 MR 指纹技术（MRF）可在单次扫描中快速且同时定量多种组织参数。目的 在体模、健康志愿者和患者中评估一种 3.0 T 快速肾脏 MRF 技术。材料与方法 设计了一个 15 秒的肾脏 MRF 采集方案，共 12 个采集段，使用了一系列低翻转角（5°-12°）、多种磁化准备方案（T1、T2 和脂肪抑制）和欠采样螺旋轨迹。该技术首先在标准化的 T1 和 T2 体模中进行了验证。随后，在 2019 年 8 月至 2020 年 10 月期间，对 10 名健康成年志愿者（平均年龄±标准差，35 岁±13；6 名男性）和 3 名常染色体隐性多囊肾病（ARPKD）的儿科患者（平均年龄，10 岁±3；2 名男性）进行了体内研究，以评估该方法的可行性。结果 在 9 个体模中的结果与基于自旋回波的 T1 和 T2 值具有良好的一致性（>0.99）。在健康成年志愿者的体内 MRF 肾脏 T1 和 T2 评估中（皮质：T1，1362 msec±5；T2，64 msec±5；髓质：T1，1827 msec±94；T2，69 msec±3），结果与文献中的值一致，但与之前的 MRF 实现相比，精度有所提高。体内 MRF 基于 B 校正的 T1 和 T2 值与无 B 校正的 T1 和 T2 值具有良好的一致性（>0.96，<0.001），表明对 B 场不均匀性的敏感性有限。使用 MRF 剖面的前 9 个段（11 秒采集时间）进行的其他 MRF 重建与使用 12 个段（15 秒采集时间）进行的重建具有良好的一致性（>0.87，<0.001）。对 3 名 ARPKD 患者连续数天进行的重复肾脏 MRF 扫描也显示出良好的可重复性（T1 和 T2：<3%的差异）。结论 在单次屏气中，一种肾脏 MRF 方法可在 3.0 T 下提供体内肾脏 T1 和 T2 图，具有更高的精度，且无需 B 校正。
