3.0T水脂分离梯度回波成像定量分析肝脏脂肪含量的可行性

目的 探讨3.0T水脂分离梯度回波序列在一次屏息内完成定量分析肝脏脂肪的可行性及其参数优化。方法 利用GE Signa HDx 3.0T MR系统对自制体模及42名受检者行迭代最小二乘法非对称采集水脂分离（IDEAL）梯度回波序列和单体素¹H-MRS，通过分析接收带宽（BW）、翻转角（FA）和矩阵（Matrix）的变化对体模脂肪定量分析的影响，确定以IDEAL梯度回波行脂肪定量分析的最优成像参数；以¹H-MRS测定的脂肪含量为参考标准，采用Spearman相关分析对IDEAL梯度回波序列测得数据进行相关性分析。结果 应用IDEAL梯度回波序列行体模脂肪定量分析时，BW和FA对结果的影响较大；BW为200 kHz、FA为12°时，体模脂肪定量分析结果与MRS的关联性最佳（r=0.997，P<0.05），在此条件下，应用IDEAL梯度回波序列和¹H-MRS测得的肝脏脂肪含量分别为（9.48±5.42）%和（10.13±8.06）%，二者呈正相关（r=0.872，P<0.05）。结论 经参数优化的IDEAL梯度回波序列可在一次屏息内定量分析肝脏脂肪含量；成像时需合理调整FA和BW。

Feasibility of liver fat quantification using IDEAL gradient echo imaging at 3.0T MR

Objective To observe the feasibility of liver fat quantification using iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL) gradient echo imaging at 3.0T MR system, and to optimize its imaging parameters. Methods IDEAL gradient echo imaging and single voxel ¹H-MRS of phantoms and 42 healthy volunteers were performed on a GE Signa HDx 3.0T MR system. Fat quantification deviations caused by variation of flip angle, bandwidth and matrix were analyzed to optimize IDEAL gradient echo imaging parameters. Taking ¹H-MRS as reference standards, the accuracy of liver fat fraction obtained by parameter-optimized IDEAL gradient echo imaging was evaluated using Spearman statistical method. Results Fat quantification results varied with the changes of bandwidth and flip angle. The best results of liver fat fraction were achieved under the condition of flip angle 12°, along with bandwidth 200 kHz, and the correlation of phantoms fat quantification between IDEAL gradient echo imaging and MRS was very good (r=0.997, P<0.05). The mean liver fat fraction of 42 subjects obtained with MR imaging and MRS was (9.48±5.42)% and (10.13±8.06)% ( r=0.872,P<0.05). Conclusion IDEAL gradient echo imaging with the optimized parameters is feasible to apply in clinical for quantification of liver fat fraction during one breath, with a fine tuning of flip angle and bandwidth.