3.0T磁共振大鼠肾脏血氧水平依赖成像的初步研究

目的：评价用3．0TMR行大鼠肾脏功能成像的可行性，评价R2^＊值在观察者问和观察者内的可重复性。方法：12只健康雄性Wistar大鼠分别行肾脏T2WI及BOLD成像，两位医师分别对T2WI及BOLD图像按照5分法进行质量评价。在T2图像上由一位医师分别测量左右肾皮髓质信噪比。使用工作站进行BOLD图像重建，由两位医师分别独立测量肾皮髓质R2^＊值，其中一位医师2周后再次测量。然后对测量及评价结果进行统计学处理。结果：两观察者分别认为T2WI有91．7％（11／12）和83．3％（10／12）达到5分（Kappa值0．625，P=0．02）；BOLD图像两者判断分别有83．3％（10／12）和91．7％（11／12）达到4分及以上（Kappa值0．586，P=0．008）均达到测量要求。Bland—Altman分析显示观察者问和观察者自身测量皮髓质R2^＊值重复性较好，相关系数分别达到观察者问：0．96，0．98；观察者自身：0．97，0．98。R2^＊值的变异系数为皮质17．80％；髓质17．96％。结论：3．0TMR行大鼠肾脏BOLD成像所测数据在不同观察者问和同一观察者内的可重复性较好，能够用来进行肾脏功能成像的实验研究。

Blood oxygen level-dependent MR imaging of rat kidney at 3.0T MR scanner: initial experience

Objective To evaluate the feasibility of BOLD MR imaging of rat kidney at 3.0T MR scanner and the interobserver and intraobserver reproducibility of R2 value. Methods Twelve male healthy Wistar rats were performed kidney T2WI and BOLD imaging. Two doctors evaluated the quality of the T2WI and BOLD images with 5 grades criterion. Bilateral kidney SNRs of cortex and medulla were measured on T2WI. BOLD images were reconstructed on workstation, two doctors measured the R2 values of cortex and medulla respectively and one of the two doctors measured them again 2 weeks later. The values were statistically analyzed. Results 91.7%(11/12) vs 83.3%(10/12) of T2W images scored 5 grades (Kappa value=0.625, P=0.02), and 83.3% (10/12) vs 91.7% (11/12) of BOLD images scored 4 grades and above (Kappa value=0.586,P=0.008) by the two observers respectively. Bland-Altman analysis showed that the intraobserver and interobserver reproducibility of R2 value was good. The correlation coefficient was: interobserver 0.96 and 0.98, intraobserver 0.97 and 0.98 for cortex and medulla respectively. Coefficient of variation between rats was 17.80% and 17.96% in cortex and medulla, respectively. Conclusion The interobserver and intraobserver reproducibility of R2 value of rat renal cortex and medulla measured at 3.0T MR is high, and it is feasible to perform functional MRI of rat kidney at 3.0T MR scanner.