两种场强磁共振IDEAL-IQ序列对质子密度脂肪分数和铁含量定量分析的初步对比研究

目的 比较3.0 T和1.5 T MR最小二乘法估计和不对称回波迭代分解水和脂肪成像（IDEAL-IQ）序列质子密度脂肪分数（PDFF）和铁含量（R2*）测量值的异同。 方法 于2019年12月选取佛山市第一人民医院健康志愿者20名[其中男性13名、女性7名，年龄（23.7±2.5）岁]并用不同脂肪含量的塑料试管模型进行前瞻性研究，分别在3.0 T（A组）和1.5 T（B组）2台 MRI设备上应用IDEAL-IQ序列进行扫描。在自动生成的FatFrac图和R2*图上自动勾画感兴趣区（ROI），分别测量模型、健康志愿者的肝脏和皮下脂肪的PDFF、R2*值。符合正态分布的计量资料采用配对样本t检验（方差齐）和Mann-Whitney U检验（方差不齐）进行比较。 结果 模型A1组和B1组PDFF的测量平均值分别为（20.59±14.39）%和（21.89±14.95）%，差异无统计学意义（Z=−1.550，P=0.121）；A1组和B1组R2*的测量平均值分别为（84.86±116.43） Hz和（43.61±54.59） Hz，差异有统计学意义（Z=−3.448，P=0.001）。健康志愿者3.0 T和1.5 T MRI肝脏、皮下脂肪PDFF测量平均值分别为（3.33±2.95） %和（4.39±2.80） %、（81.78±6.33） %和（81.54±5.53） %，差异均无统计学意义（Z=−1.867、t=−0.301，均P>0.05）；A组肝脏、皮下脂肪的R2*测量平均值为（52.42±12.22） Hz、（50.88±10.32） Hz，分别高于对应的B组[（32.73±5.62） Hz、（39.41±9.11） Hz]，且差异均有统计学意义（Z=−3.920、t=4.372，均P<0.001）。 结论 基于IDEAL-IQ序列的3.0 T和1.5 T MR模型、健康志愿者肝脏和皮下脂肪的PDFF测量值的差异无统计学意义，但3.0 T MRI的PDFF测量值可能比1.5 T MRI更接近临床实际情况；2种场强获得的R2*值的差异有统计学意义，其影响因素需进一步研究。

A preliminary quantitative comparative study of proton density fat fraction and iron content on two different strength of magnetic resonance with IDEAL-IQ sequence

Objective Aimed to compare similarities and differences of the quantitative measurements of liver proton density fat fraction (PDFF) and iron content (R2*) between 3.0 T and 1.5 T MRI with iterative decomposition of water and fat with asymmetry and least squares estimation quantitative fat imaging (IDEAL-IQ) sequences. Methods Twenty healthy volunteers [13 males and 7 females, aged (23.7±2.5) years] and plastic test tube models with different fat contents were scanned with IDEAL-IQ sequence for prospective research on two MRI equipment of 3.0 T (group A) and 1.5 T (group B) in December 2019 in First People's Hospital of Foshan. The FatFrac images and R2* images automatically generated were used to draw the region of interests (ROIs) automatically. The PDFF and R2* values of the models and the liver and subcutaneous fat of the healthy volunteers were measured and compared between groups A and B. Measurement data that conformed to normal distribution were compared using paired sample t test (equal variances assumed) and Mann-Whitney U test (equal variances not assumed). Results The measured mean PDFF values of the model of groups A1 and B1 were (20.59±14.39)% and (21.89±14.95)%, respectively, with no significant difference (Z=−1.550, P=0.121). The measured mean R2* values of the model of groups A1 and B1 were (84.86±116.43) Hz and (43.61±54.59) Hz, respectively, with a statistically significant difference (Z=−3.448, P=0.001). No significant difference was found in the mean PDFF values of liver and subcutaneous fat of healthy volunteers [(3.33±2.95)% vs. (4.39±2.80)%, (81.78±6.33) Hz vs. (81.54±5.53) Hz)(Z=−1.867, t=−0.301; both P>0.05] with 3.0 T and 1.5 T MRI, respectively. The average R2* measured values of group A of the liver and subcutaneous fat of the healthy volunteers were higher than those of group B, with statistical significantly differences [(52.42±12.22) Hz vs. (32.73±5.62) Hz, (50.88±10.32) Hz vs. (39.41±9.11) Hz)(Z=−3.920, t=4.372; both P<0.001]. Conclusions Results showed no significant differences in the PDFF values of models and the liver and subcutaneous fat of the volunteers measured between 3.0 T MR and 1.5 T MR based on IDEAL-IQ sequences, but the PDFF values of 3.0 T MRI may be closer to the clinical actual situation than that of 1.5 T MRI. The R2* values obtained based on 3.0 T MRI field strength were significantly different from that of 1.5 T MRI, and its influencing factors need to be further studied.